change in complex adaptive systems

42
Supporting water sanitation and hygiene services for life Change in complex adaptive systems A review of concepts, theory and approaches for tackling ‘wicked’ problems in achieving sustainable rural water services WORKING PAPER December 2015 Deirdre Casella, Simone van Tongeren, Igor Nikolic

Upload: dinhanh

Post on 10-Feb-2017

222 views

Category:

Documents


2 download

TRANSCRIPT

Page 1: Change in complex adaptive systems

Supporting water sanitation and hygiene services for life

Change in complex adaptive systems A review of concepts theory and approaches for tackling lsquowickedrsquo problems in achieving sustainable rural water services

WORKING PAPER December 2015

Deirdre Casella Simone van Tongeren Igor Nikolic

At IRC we believe that turning on a working tap should not be a surprise or cause for celebration

We believe in a world where water sanitation and hygiene services are fundamental utilities that everyone is able to take for granted For good

We face a complex challenge Every year thousands of projects within and beyond the WASH sector fail ndash the result of short-term targets and interventions at the cost of long-term service solutions

This leaves around a third of the worldrsquos poorest people without access to the most basic of human rights and leads directly to economic social and health problems on a global scale IRC exists to continually challenge and shape the established practices of the WASH sector

Through collaboration and the active application of our expertise we work with governments service providers and international organisations to deliver systems and services that are truly built to last

Change in complex adaptive systems A review of concepts theory and approaches for tackling lsquowickedrsquo problems in achieving sustainable rural water services

Prepared by Deirdre Casella MA Programme Officer IRC International Water and Sanitation Centre The Hague The Netherlands

Doctoral Candidate Faculty of Technology Policy and Management Delft Technical University The NetherlandsSimone van Tongeren MSc Management Trainee Evides Water Company Rotterdam The NetherlandsIgor Nikolic PhD Assistant Professor Faculty of Technology Policy and Management Delft Technical University The Netherlands

Key words agent-based modelling collective impact complex adaptive systems learning alliances problem-driven iterative adaption rural water services socio-technical systems systems thinking universal Darwinism whole system change

This report presents a review of theory rationale strategies and methods that underpin approaches to effecting lasting change in complex adaptive systems The context is the water sanitation and hygiene (WASH) services sector in developing countries The report discusses the multi-faceted challenges that the sector faces in delivering sustainable services and summarises the growing calls for a complexity-informed transformation in how the development aid system is conceptualised and how its products are delivered Whole system change approaches are presented and discussed for their potential to achieve this transformation Reviewers Catarina Fonseca PhD and Patrick Moriarty PhD

Edited by Sally Atwater and laid out by Ghislaine Heylen For questions or clarifications contact IRC here wwwircwashorgcontact-us

Design and layout Punt Grafisch Ontwerp

Deirdre Casella Simone van Tongeren Igor Nikolic

Change in complex adaptive systems A review of concepts theory and approaches for tackling lsquowickedrsquo problems in achieving sustainable rural water services

Contents

ABBREVIATIONS 4

1 PURPOSE OF THIS REVIEW 5

2 WATER SERVICES A SYSTEM DESCRIPTION 721 Rural water services ndash from governance to systems perspectives 722 International development aid Exogenous change pressure 923 Capital investments but failing services 1224 Calls to re-think international development aid 12

3 CONCEPTS FROM THE COMPLEXITY SCIENCES 1431 Complex adaptive systems and the socio-technical system 1432 Change in socio-technical systems 1533 Universal Darwinism A meta-theory of evolving systems 1734 The language of institutions IAD and ADICO 1835 Rural water services as a socio-technical system 19

4 APPROACHES TO WHOLE SYSTEM CHANGE 2241 Collective impact Creating large-scale social change 2242 Learning alliances 2343 Problem-driven iterative adaptation 25

5 SIMULATING COMPLEX PROBLEMS TOOLS TO TEST AND LEARN 26 51 AGENT-BASED MODELLING 2652 SERIOUS GAMES 28

6 DISCUSSION AND CONCLUSION 30

REFERENCES 32

TablesTable 1 Agent-based model conceptualisation Rules of local government 29

FiguresFigure 1 Development partner funding and national allocations to WASH as percentage of GDP 10Figure 2 Economics of institutions 11Figure 3 IAD framework 19Figure 4 Republic of Uganda rural water sector actor landscape 21Figure 5 Sector change and innovation process 23Figure 6 Learning alliance approach to scaling change across institutional levels 24Figure 7 Structure of an agent-based model 27

4

Abbreviations

ADIC ADICO AIC CAS CBO GDP IAD NGO OECD PDIA SOHO WASH

attributes deontic aim conditionattributes deontic aim condition or elseattributes aim conditioncomplex adaptive systemcommunity-based organisationgross domestic productinstitutional analysis and developmentnon-governmental organisationOrganisation for Economic Co-operation and Developmentproblem-driven iterative adaptationself-organising hierarchical openwater sanitation and hygiene

5

1 Purpose of this review

This working paper a product of the Triple-S Water Services That Last project 1 is part of an effort to ground IRCrsquos approach to driving and supporting change processes in the water sanitation and hygiene (WASH) sector in the theory and methods of complexity sciences and systems thinking The review of literature from these fields identifies theory rationale and methods underpinning complexity-informed approaches to effecting change in large dynamic complex adaptive systems 2 The insights gained from this body of literature are discussed in relation to the rural water sector in low- and middle-income countriesmdasha sector facing lsquowickedrsquo 3 problems whose solution requires changes in the mindset and behaviours of multiple actors (Rittel and Webber 1973)

The domain of inquiry of this review is the rural water sector in low- and middle-income countries Section 2 describes a series of generally recognisable attributes of the rural water sector as they feature in low- and middle-income countries Salient attributes include the roles and responsibilities of key actors and patterns of interaction between and among key actors institutions and technology in this domain This section also reflects on the multi-faceted and inextricably linked challenges to delivering sustainable services and on the role of international development aid as an external influence in national development agendas and thereby in how public services are financed planned and implemented It reviews the literature on how international development aid can address these challenges by embracing concepts and practices informed by systems thinking and complexity sciences

Section 3 presents concepts and approaches from the complexity sciences and related fields including socio-technical systems as a specific type of complex adaptive system institutional analysis and whole-system change and explores whether they offer new perspectives on alternatives to prevailing national development and water service delivery policy and practice in particular Following the reflection on insights offered by the complexity sciences Section 3 concludes by reconsidering the rural water sector in light of the theory about complex adaptive systems and how change arises in such systems

Section 4 discusses three approaches featured in the literature for their potential to effect systemic change in complex adaptive systems as proposed alternatives to current policy making and planning in international development Given that large-scale systemic change can take several decades Section 5 presents tools for simulating change processes in complex adaptive systems for the purpose of envisaging exploring and experimenting with policy and implementation alternatives Section 6 concludes with a discussion of the insights gained through this review and identifies gaps in the literature about how to foster national systems that can deliver sustainable public services

1 Triple-S Water Services That Last is a six-year multi-country learning initiative to improve water supply to the rural poor that is led by IRC a Netherlands-based mission driven lsquothink and do tankrsquo pursuing the vision of WASH services for everyone forever The initiative is funded by the Bill amp Melinda Gates Foundation and has country programmes active in Ghana and Uganda with smaller initiatives supported in Burkina Faso Mozambique India and Honduras Lessons learned from work in countries feeds up to the international level where Triple-S promotes a re-appraisal of how development assistance to the rural water supply sector is designed and implemented

2 Moriarty Lockwood Carriger and Duti series of four blogposts httpwaterservicesthatlastwordpresscom20140224changing-the-whole-system-to-provide-water-sanitation-and-hygiene-services-that-last MarchndashMay 2014

3 lsquoWickedrsquo describes intractable not-easily-solved problemsmdashsuch as climate change pandemics poverty or natural disastersmdashthat require changes in mindset and behaviours of a great number of actors (Rittel and Webber 1973)

6

WORKING PAPER Change in complex adaptive systems

7

December 2015

In this section the rural water sub-sector and the challenges it faces in providing sustainable services are described Section 21 looks at the administrative and operational arrangements as well as roles and functions of actors involved in rural water sector in low- and middle-income countries Section 22 presents literature about the role and influence of international development as an external change pressure that influences domestic policy financial and technical aspects of rural water services in low- and middle-income countries Section 23 provides an overview of the multiple interlinked and intractable lsquowickedrsquo challenges to delivering services reflecting on the fact that the solutions are not hardware based Section 24 reviews literature arguing that such wicked challenges require alternative approaches grounded in the complexity sciences

21 RURAL WATER SERVICES ndash FROM GOVERNANCE TO SYSTEMS PERSPECTIVES

The overarching purpose of a national WASH sector is to develop and deliver sustainable water sanitation and hygiene services to users for domestic and productive purposes (Plummer and Slaymaker 2007) Although the WASH sector is not clearly demarcated as an entity with precisely identifiable boundaries a number of salient attributes can be identified across different national and regional contexts that enable us to consider the sector a lsquosystemrsquo There are also a number of identifiable activities that people must undertake to ensure that water services are delivered These attributes activities and interactions taken together give rise to a recognisable pattern that when effectively functioning results in the delivery of water services In this section the rural water sector and its salient attributes are described as the domain of focus of this literature review

In terms of attributes there are identifiable legal governance and organisational arrangements in place within nation states that guide the delivery of water services (Rogers and Hall 2003) In general a line ministry or department is the highest mandated authority responsible for ensuring that these services are delivered to all citizens In addition to developing and managing the natural water resource this national authority is tasked with developing and delivering water services to the population for domestic and productive purposes While the national authority for water resources management and

service delivery may also hold the remit for development and delivery of sanitation and hygiene services for the purpose of this review we focus in particular on the legal and administrative arrangements pertaining to rural water service delivery

In the context of decentralised water service delivery models in addition to the national government many other organisations and actors interacting on a continual basis are involved in policy making financial planning and management regulation and service provision activities across multiple administrative levels (Rogers and Hall 2003) Depending upon the national context these actors include sub-national government entities (eg provinces regions districts communes zones municipalities woredas panchayats) water utilities non-governmental organisations (NGOs) users community representatives private operators and capacity-building and financing bodies (de la Harpe 2007) Among this host of actors roles and responsibilities can be differentiated As highlighted by Smits et al (2011) a key distinction is the role of the service authority versus that of the service provider

The service authority generally a government body holds the legal responsibility for service delivery planning coordination regulation and oversight activities as well as technical assistance to water service providers (Smits et al 2011 Lockwood and Smits 2011) In contrast the service provider is the organisation or individual responsible for day-to-day water service which includes the operation maintenance and administration of the water system

How the water service provision role is fulfilled varies widely Most countries have a range of service provision options or service delivery models whereby lsquothe service authority can opt to provide services itself (through a municipal department or municipal company) or hellip delegate this responsibility by contracting an outside agency such as a community-based organisation (CBO) private operator public sector utility or company or non-governmental organisation (NGO) who in turn may hire a private person (plumber or mechanic) to carry out parts of the workrsquo (Smits et al 2011 p5)

In addition to the legal administrative and organisational attributes and interactionsmdashthat is the lsquosocialrsquo components of a water servicemdashthere is also

2 Water services A system description

8

WORKING PAPER Change in complex adaptive systems

the technical component The physical infrastructure required for the delivery of water supply varies greatly among and even within service areas4 depending upon factors such as geographical and hydrological conditions preferences for certain technologies available financial resources and population size and density

Whether the infrastructure is a stand-alone hand pump or a networked gravity-fed piped scheme formal and informal arrangements among the authority provider users civil society and international development organisations are required to ensure sustainable water services (Keohane and Ostrom 1995 Rogers and Hall 2003 Plummer and Slaymaker 2007) These arrangements entail policy- and decision-making processes about responsibilities and actor relationships through which the power responsibilities norms values and formal agreements embedded in laws and policies are negotiated among and implemented by the array of stakeholders whose roles and responsibilities may overlap (Ostrom and Janssen 2004)

This view of interdependent networks of multiple actors or agents interacting across multiple administrative levels embodies a lsquogovernance perspectiversquo (Kooiman 1993 Stoker 1998 DFID 2007) In reflecting on how to provide for a collective interest such as the provision of public services the governance perspective offers a lsquoframework for understanding changing processes of governing characterised by processes of adaptation learning and experimentrsquo (Stoker 1998 p18) This perspective helps to conceptualise how functions related to service delivery are dispersed over a wide array of actors organisations and coordination platforms spanning different national development sectors and administrative levels Notably from a governance perspective while (central) governments continue to play a role in how public services are provided as Bache (2003) notes this role is increasingly one of coordination and steering and is concomitant with an increase in the involvement of non-government actors in policy-making and service delivery The Global Water Partnership (GWP) defines water

governance as lsquothe range of political social economic and administrative systems that are in place to develop and manage water resources and the delivery of water services at different levels of societyrsquo (Rogers and Hall 2003) The literature on governance also highlights the importance of public participation in governance processes for the potential to lsquoimprove the quality of decision making by opening up the decision-making process and making better use of the information and creativity that is available in society improve public under-standing of the management issues at stake make decision making more transparent and might stimulate the different government bodies involved to coordinate their actions more in order to provide serious follow-up to the inputs receivedrsquo and potentially strengthen democratic processes where government does not have all the resources required to lsquomanage an issue effectivelyrsquo (Huitema et al 2009 p5)

The prevailing governance approaches in low- and middle-income countries are context specific and have evolved over long periods of time in response to change pressures such as political social and economic processes (Plummer and Slaymaker 2007) Irrespective of the governance arrangements in a given national context the governance perspective as described here makes it possible to introduce a systems perspective5 to understanding the rural water sector6 This line of inquiry will be addressed in further detail in Section 3

By framing water services as a system that is open to feedback from its environment it becomes possible to consider external change pressures increasingly recognised for their effect on how national development agendasmdashincluding water service deliverymdashare formed and implemented One such pressure is the policy and finance support provided to low- and middle-income countries in the form of international development aid from parties such as UN agencies international finance institutes philanthropic organisations non-governmental organisations and middle- and high-income countries (Rogers and Hall 2003 Mowles et al 2008

4 A service area is the area of jurisdiction and population covered by a service authority Service areas are typically linked to the boundaries of human settlement (towns villages hamlets and scattered rural settlements) but may not correspond precisely with administrative boundaries (IRC Glossary accessed 16 October 2014)

5 A systems perspective takes into account all of the behaviours of a system as a whole in the context of its environment and is lsquoa non-reductionist approach to describing the properties of the system itselfrsquo A description of the whole must include an explanation of the relationships between the parts as well as any additional information needed to describe the behaviour of the entire system (after Bar-Yam 1997 2005)

6 A system is defined by Ryan (2008) as lsquoa representation of an entity as a complex whole open to feedback from its environmentrsquo Ryan (2008) Burke (2006) and other authors on systems and complexity sciences make the important observation that such representations are idealisations based on simplified assumptions Thus although they offer a valid means for identifying and analysing an entity and its dynamics lsquothere are limits to their applicationrsquo (Mowles et al 2008 Ryan 2008)

9

December 2015

Ramalingam 2013) Given the importance of international development aid in shaping national development agenda priorities the next section summarises literature discussing its role and reflects on the dynamic arising in the water sector where it is a main source of finance for service development and a highly influential exogenous change pressure

22 INTERNATIONAL DEVELOPMENT AID EXOGENOUS CHANGE PRESSURE

In this sub-section the role of international development aid in shaping national development agendas of low- and middle-income countries is discussed The international development aid system provides policy financial and technical support to developing nations for education health transportation energy and local and regional economic development and trade A clear understanding of the dynamic created by the involvement of such influential external agents is critical to understanding the current challenges to achieving sustainable water services let alone the realisation of resilient national systems that can develop and deliver the public services required for sustainable and equitable social and economic development (Mowles et al 2008) Many researchers have explored how the current architecture of international development aid delivery is hindering the potential of nations to achieve these social and economic development goals

When developing countries need capital to build infrastructure for public services international development aid actorsmdashranging from development banks funders and bi-lateral government agencies to NGOs and philanthropic organisations (henceforth lsquodevelopment partnersrsquo)mdashmay explicitly partner with a recipient government through policy budget andor technical support strategies and implementation plans (Mowles et al 2008 Ramalingam 2013) It is not uncommon for development partners to bypass nationally led processes and directly implement programmes at the user and community levels (Nimanya et al 2011)

The visions missions and mandates of development partners vary greatly and determine the nature of their interactions with domestic partners including financial investment decisions In addition to investing in infrastructure development partners may support organisational policy and capacity aspects related to the sustainable delivery of a service including both direct support (monitoring maintenance repairs replacements training of staff)

and indirect support (macro-level planning and policy making) These essential components of sustainable water services known as post-construction support or lsquosoftwarersquo are as important as the infrastructure lsquohardwarersquo yet often neglected with actual levels of financial support considered insufficient (Rogers and Hall 2003 Smits et al 2011)

Reasons for this neglect include the desire to focus resources on increasing coverage rates for unserved populations (WHO 2012) perceptions about the risk of corruption faulty assumptions about the lsquobestrsquo governance arrangements for post-construction activities (Schouten and Moriarty 2003) and the desire to see tangible easily measurable results from an investment (Garandeau et al 2009)

Understanding the sources of financing provides insight into how WASH policy priorities are determined The Organisation for Economic Co-operation and Development (OECD 2009) distinguishes the sources of financing for WASH services as the three Ts tariffs which are contributions paid by service users transfers in the form of assistance from development partners and taxes which are levied by national or regional governments In many countries the funding from international aid (transfers to use the OECD nomenclature) is at least as much as the funding from the two domestic sources tariffs and taxes (Figure 1)

Financing from transfers is not problematic on its own and in fact does tremendous good in many countries However since lsquohe who pays the piper calls the tunersquo national policy strategy and governance reform interventions are frequently and significantly influenced by development partnersrsquo priorities (Water Aid 2011) especially where transfers are collectively greater than domestic sources of financing generated through taxes and tariffs and where transfers are made outside the national policy agenda

As Figure 1 shows lsquodonor aid to the WASH sector as a percentage of GDP is higher than government budget allocations for WASH in Cambodia Ghana Liberia Madagascar Rwanda Timor-Leste and Uganda indicating both a donor-dominated sector and also that significant amounts of aid to the WASH sector in these countries is not recorded in central government budgets and accounts or is off-budgetrsquo (WaterAid 2011 p35) This disparity in international and domestic funding translates into disproportionate levels of influence by development partners in shaping national and sub-national development agenda priorities (WaterAid 2011)

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 2: Change in complex adaptive systems

At IRC we believe that turning on a working tap should not be a surprise or cause for celebration

We believe in a world where water sanitation and hygiene services are fundamental utilities that everyone is able to take for granted For good

We face a complex challenge Every year thousands of projects within and beyond the WASH sector fail ndash the result of short-term targets and interventions at the cost of long-term service solutions

This leaves around a third of the worldrsquos poorest people without access to the most basic of human rights and leads directly to economic social and health problems on a global scale IRC exists to continually challenge and shape the established practices of the WASH sector

Through collaboration and the active application of our expertise we work with governments service providers and international organisations to deliver systems and services that are truly built to last

Change in complex adaptive systems A review of concepts theory and approaches for tackling lsquowickedrsquo problems in achieving sustainable rural water services

Prepared by Deirdre Casella MA Programme Officer IRC International Water and Sanitation Centre The Hague The Netherlands

Doctoral Candidate Faculty of Technology Policy and Management Delft Technical University The NetherlandsSimone van Tongeren MSc Management Trainee Evides Water Company Rotterdam The NetherlandsIgor Nikolic PhD Assistant Professor Faculty of Technology Policy and Management Delft Technical University The Netherlands

Key words agent-based modelling collective impact complex adaptive systems learning alliances problem-driven iterative adaption rural water services socio-technical systems systems thinking universal Darwinism whole system change

This report presents a review of theory rationale strategies and methods that underpin approaches to effecting lasting change in complex adaptive systems The context is the water sanitation and hygiene (WASH) services sector in developing countries The report discusses the multi-faceted challenges that the sector faces in delivering sustainable services and summarises the growing calls for a complexity-informed transformation in how the development aid system is conceptualised and how its products are delivered Whole system change approaches are presented and discussed for their potential to achieve this transformation Reviewers Catarina Fonseca PhD and Patrick Moriarty PhD

Edited by Sally Atwater and laid out by Ghislaine Heylen For questions or clarifications contact IRC here wwwircwashorgcontact-us

Design and layout Punt Grafisch Ontwerp

Deirdre Casella Simone van Tongeren Igor Nikolic

Change in complex adaptive systems A review of concepts theory and approaches for tackling lsquowickedrsquo problems in achieving sustainable rural water services

Contents

ABBREVIATIONS 4

1 PURPOSE OF THIS REVIEW 5

2 WATER SERVICES A SYSTEM DESCRIPTION 721 Rural water services ndash from governance to systems perspectives 722 International development aid Exogenous change pressure 923 Capital investments but failing services 1224 Calls to re-think international development aid 12

3 CONCEPTS FROM THE COMPLEXITY SCIENCES 1431 Complex adaptive systems and the socio-technical system 1432 Change in socio-technical systems 1533 Universal Darwinism A meta-theory of evolving systems 1734 The language of institutions IAD and ADICO 1835 Rural water services as a socio-technical system 19

4 APPROACHES TO WHOLE SYSTEM CHANGE 2241 Collective impact Creating large-scale social change 2242 Learning alliances 2343 Problem-driven iterative adaptation 25

5 SIMULATING COMPLEX PROBLEMS TOOLS TO TEST AND LEARN 26 51 AGENT-BASED MODELLING 2652 SERIOUS GAMES 28

6 DISCUSSION AND CONCLUSION 30

REFERENCES 32

TablesTable 1 Agent-based model conceptualisation Rules of local government 29

FiguresFigure 1 Development partner funding and national allocations to WASH as percentage of GDP 10Figure 2 Economics of institutions 11Figure 3 IAD framework 19Figure 4 Republic of Uganda rural water sector actor landscape 21Figure 5 Sector change and innovation process 23Figure 6 Learning alliance approach to scaling change across institutional levels 24Figure 7 Structure of an agent-based model 27

4

Abbreviations

ADIC ADICO AIC CAS CBO GDP IAD NGO OECD PDIA SOHO WASH

attributes deontic aim conditionattributes deontic aim condition or elseattributes aim conditioncomplex adaptive systemcommunity-based organisationgross domestic productinstitutional analysis and developmentnon-governmental organisationOrganisation for Economic Co-operation and Developmentproblem-driven iterative adaptationself-organising hierarchical openwater sanitation and hygiene

5

1 Purpose of this review

This working paper a product of the Triple-S Water Services That Last project 1 is part of an effort to ground IRCrsquos approach to driving and supporting change processes in the water sanitation and hygiene (WASH) sector in the theory and methods of complexity sciences and systems thinking The review of literature from these fields identifies theory rationale and methods underpinning complexity-informed approaches to effecting change in large dynamic complex adaptive systems 2 The insights gained from this body of literature are discussed in relation to the rural water sector in low- and middle-income countriesmdasha sector facing lsquowickedrsquo 3 problems whose solution requires changes in the mindset and behaviours of multiple actors (Rittel and Webber 1973)

The domain of inquiry of this review is the rural water sector in low- and middle-income countries Section 2 describes a series of generally recognisable attributes of the rural water sector as they feature in low- and middle-income countries Salient attributes include the roles and responsibilities of key actors and patterns of interaction between and among key actors institutions and technology in this domain This section also reflects on the multi-faceted and inextricably linked challenges to delivering sustainable services and on the role of international development aid as an external influence in national development agendas and thereby in how public services are financed planned and implemented It reviews the literature on how international development aid can address these challenges by embracing concepts and practices informed by systems thinking and complexity sciences

Section 3 presents concepts and approaches from the complexity sciences and related fields including socio-technical systems as a specific type of complex adaptive system institutional analysis and whole-system change and explores whether they offer new perspectives on alternatives to prevailing national development and water service delivery policy and practice in particular Following the reflection on insights offered by the complexity sciences Section 3 concludes by reconsidering the rural water sector in light of the theory about complex adaptive systems and how change arises in such systems

Section 4 discusses three approaches featured in the literature for their potential to effect systemic change in complex adaptive systems as proposed alternatives to current policy making and planning in international development Given that large-scale systemic change can take several decades Section 5 presents tools for simulating change processes in complex adaptive systems for the purpose of envisaging exploring and experimenting with policy and implementation alternatives Section 6 concludes with a discussion of the insights gained through this review and identifies gaps in the literature about how to foster national systems that can deliver sustainable public services

1 Triple-S Water Services That Last is a six-year multi-country learning initiative to improve water supply to the rural poor that is led by IRC a Netherlands-based mission driven lsquothink and do tankrsquo pursuing the vision of WASH services for everyone forever The initiative is funded by the Bill amp Melinda Gates Foundation and has country programmes active in Ghana and Uganda with smaller initiatives supported in Burkina Faso Mozambique India and Honduras Lessons learned from work in countries feeds up to the international level where Triple-S promotes a re-appraisal of how development assistance to the rural water supply sector is designed and implemented

2 Moriarty Lockwood Carriger and Duti series of four blogposts httpwaterservicesthatlastwordpresscom20140224changing-the-whole-system-to-provide-water-sanitation-and-hygiene-services-that-last MarchndashMay 2014

3 lsquoWickedrsquo describes intractable not-easily-solved problemsmdashsuch as climate change pandemics poverty or natural disastersmdashthat require changes in mindset and behaviours of a great number of actors (Rittel and Webber 1973)

6

WORKING PAPER Change in complex adaptive systems

7

December 2015

In this section the rural water sub-sector and the challenges it faces in providing sustainable services are described Section 21 looks at the administrative and operational arrangements as well as roles and functions of actors involved in rural water sector in low- and middle-income countries Section 22 presents literature about the role and influence of international development as an external change pressure that influences domestic policy financial and technical aspects of rural water services in low- and middle-income countries Section 23 provides an overview of the multiple interlinked and intractable lsquowickedrsquo challenges to delivering services reflecting on the fact that the solutions are not hardware based Section 24 reviews literature arguing that such wicked challenges require alternative approaches grounded in the complexity sciences

21 RURAL WATER SERVICES ndash FROM GOVERNANCE TO SYSTEMS PERSPECTIVES

The overarching purpose of a national WASH sector is to develop and deliver sustainable water sanitation and hygiene services to users for domestic and productive purposes (Plummer and Slaymaker 2007) Although the WASH sector is not clearly demarcated as an entity with precisely identifiable boundaries a number of salient attributes can be identified across different national and regional contexts that enable us to consider the sector a lsquosystemrsquo There are also a number of identifiable activities that people must undertake to ensure that water services are delivered These attributes activities and interactions taken together give rise to a recognisable pattern that when effectively functioning results in the delivery of water services In this section the rural water sector and its salient attributes are described as the domain of focus of this literature review

In terms of attributes there are identifiable legal governance and organisational arrangements in place within nation states that guide the delivery of water services (Rogers and Hall 2003) In general a line ministry or department is the highest mandated authority responsible for ensuring that these services are delivered to all citizens In addition to developing and managing the natural water resource this national authority is tasked with developing and delivering water services to the population for domestic and productive purposes While the national authority for water resources management and

service delivery may also hold the remit for development and delivery of sanitation and hygiene services for the purpose of this review we focus in particular on the legal and administrative arrangements pertaining to rural water service delivery

In the context of decentralised water service delivery models in addition to the national government many other organisations and actors interacting on a continual basis are involved in policy making financial planning and management regulation and service provision activities across multiple administrative levels (Rogers and Hall 2003) Depending upon the national context these actors include sub-national government entities (eg provinces regions districts communes zones municipalities woredas panchayats) water utilities non-governmental organisations (NGOs) users community representatives private operators and capacity-building and financing bodies (de la Harpe 2007) Among this host of actors roles and responsibilities can be differentiated As highlighted by Smits et al (2011) a key distinction is the role of the service authority versus that of the service provider

The service authority generally a government body holds the legal responsibility for service delivery planning coordination regulation and oversight activities as well as technical assistance to water service providers (Smits et al 2011 Lockwood and Smits 2011) In contrast the service provider is the organisation or individual responsible for day-to-day water service which includes the operation maintenance and administration of the water system

How the water service provision role is fulfilled varies widely Most countries have a range of service provision options or service delivery models whereby lsquothe service authority can opt to provide services itself (through a municipal department or municipal company) or hellip delegate this responsibility by contracting an outside agency such as a community-based organisation (CBO) private operator public sector utility or company or non-governmental organisation (NGO) who in turn may hire a private person (plumber or mechanic) to carry out parts of the workrsquo (Smits et al 2011 p5)

In addition to the legal administrative and organisational attributes and interactionsmdashthat is the lsquosocialrsquo components of a water servicemdashthere is also

2 Water services A system description

8

WORKING PAPER Change in complex adaptive systems

the technical component The physical infrastructure required for the delivery of water supply varies greatly among and even within service areas4 depending upon factors such as geographical and hydrological conditions preferences for certain technologies available financial resources and population size and density

Whether the infrastructure is a stand-alone hand pump or a networked gravity-fed piped scheme formal and informal arrangements among the authority provider users civil society and international development organisations are required to ensure sustainable water services (Keohane and Ostrom 1995 Rogers and Hall 2003 Plummer and Slaymaker 2007) These arrangements entail policy- and decision-making processes about responsibilities and actor relationships through which the power responsibilities norms values and formal agreements embedded in laws and policies are negotiated among and implemented by the array of stakeholders whose roles and responsibilities may overlap (Ostrom and Janssen 2004)

This view of interdependent networks of multiple actors or agents interacting across multiple administrative levels embodies a lsquogovernance perspectiversquo (Kooiman 1993 Stoker 1998 DFID 2007) In reflecting on how to provide for a collective interest such as the provision of public services the governance perspective offers a lsquoframework for understanding changing processes of governing characterised by processes of adaptation learning and experimentrsquo (Stoker 1998 p18) This perspective helps to conceptualise how functions related to service delivery are dispersed over a wide array of actors organisations and coordination platforms spanning different national development sectors and administrative levels Notably from a governance perspective while (central) governments continue to play a role in how public services are provided as Bache (2003) notes this role is increasingly one of coordination and steering and is concomitant with an increase in the involvement of non-government actors in policy-making and service delivery The Global Water Partnership (GWP) defines water

governance as lsquothe range of political social economic and administrative systems that are in place to develop and manage water resources and the delivery of water services at different levels of societyrsquo (Rogers and Hall 2003) The literature on governance also highlights the importance of public participation in governance processes for the potential to lsquoimprove the quality of decision making by opening up the decision-making process and making better use of the information and creativity that is available in society improve public under-standing of the management issues at stake make decision making more transparent and might stimulate the different government bodies involved to coordinate their actions more in order to provide serious follow-up to the inputs receivedrsquo and potentially strengthen democratic processes where government does not have all the resources required to lsquomanage an issue effectivelyrsquo (Huitema et al 2009 p5)

The prevailing governance approaches in low- and middle-income countries are context specific and have evolved over long periods of time in response to change pressures such as political social and economic processes (Plummer and Slaymaker 2007) Irrespective of the governance arrangements in a given national context the governance perspective as described here makes it possible to introduce a systems perspective5 to understanding the rural water sector6 This line of inquiry will be addressed in further detail in Section 3

By framing water services as a system that is open to feedback from its environment it becomes possible to consider external change pressures increasingly recognised for their effect on how national development agendasmdashincluding water service deliverymdashare formed and implemented One such pressure is the policy and finance support provided to low- and middle-income countries in the form of international development aid from parties such as UN agencies international finance institutes philanthropic organisations non-governmental organisations and middle- and high-income countries (Rogers and Hall 2003 Mowles et al 2008

4 A service area is the area of jurisdiction and population covered by a service authority Service areas are typically linked to the boundaries of human settlement (towns villages hamlets and scattered rural settlements) but may not correspond precisely with administrative boundaries (IRC Glossary accessed 16 October 2014)

5 A systems perspective takes into account all of the behaviours of a system as a whole in the context of its environment and is lsquoa non-reductionist approach to describing the properties of the system itselfrsquo A description of the whole must include an explanation of the relationships between the parts as well as any additional information needed to describe the behaviour of the entire system (after Bar-Yam 1997 2005)

6 A system is defined by Ryan (2008) as lsquoa representation of an entity as a complex whole open to feedback from its environmentrsquo Ryan (2008) Burke (2006) and other authors on systems and complexity sciences make the important observation that such representations are idealisations based on simplified assumptions Thus although they offer a valid means for identifying and analysing an entity and its dynamics lsquothere are limits to their applicationrsquo (Mowles et al 2008 Ryan 2008)

9

December 2015

Ramalingam 2013) Given the importance of international development aid in shaping national development agenda priorities the next section summarises literature discussing its role and reflects on the dynamic arising in the water sector where it is a main source of finance for service development and a highly influential exogenous change pressure

22 INTERNATIONAL DEVELOPMENT AID EXOGENOUS CHANGE PRESSURE

In this sub-section the role of international development aid in shaping national development agendas of low- and middle-income countries is discussed The international development aid system provides policy financial and technical support to developing nations for education health transportation energy and local and regional economic development and trade A clear understanding of the dynamic created by the involvement of such influential external agents is critical to understanding the current challenges to achieving sustainable water services let alone the realisation of resilient national systems that can develop and deliver the public services required for sustainable and equitable social and economic development (Mowles et al 2008) Many researchers have explored how the current architecture of international development aid delivery is hindering the potential of nations to achieve these social and economic development goals

When developing countries need capital to build infrastructure for public services international development aid actorsmdashranging from development banks funders and bi-lateral government agencies to NGOs and philanthropic organisations (henceforth lsquodevelopment partnersrsquo)mdashmay explicitly partner with a recipient government through policy budget andor technical support strategies and implementation plans (Mowles et al 2008 Ramalingam 2013) It is not uncommon for development partners to bypass nationally led processes and directly implement programmes at the user and community levels (Nimanya et al 2011)

The visions missions and mandates of development partners vary greatly and determine the nature of their interactions with domestic partners including financial investment decisions In addition to investing in infrastructure development partners may support organisational policy and capacity aspects related to the sustainable delivery of a service including both direct support (monitoring maintenance repairs replacements training of staff)

and indirect support (macro-level planning and policy making) These essential components of sustainable water services known as post-construction support or lsquosoftwarersquo are as important as the infrastructure lsquohardwarersquo yet often neglected with actual levels of financial support considered insufficient (Rogers and Hall 2003 Smits et al 2011)

Reasons for this neglect include the desire to focus resources on increasing coverage rates for unserved populations (WHO 2012) perceptions about the risk of corruption faulty assumptions about the lsquobestrsquo governance arrangements for post-construction activities (Schouten and Moriarty 2003) and the desire to see tangible easily measurable results from an investment (Garandeau et al 2009)

Understanding the sources of financing provides insight into how WASH policy priorities are determined The Organisation for Economic Co-operation and Development (OECD 2009) distinguishes the sources of financing for WASH services as the three Ts tariffs which are contributions paid by service users transfers in the form of assistance from development partners and taxes which are levied by national or regional governments In many countries the funding from international aid (transfers to use the OECD nomenclature) is at least as much as the funding from the two domestic sources tariffs and taxes (Figure 1)

Financing from transfers is not problematic on its own and in fact does tremendous good in many countries However since lsquohe who pays the piper calls the tunersquo national policy strategy and governance reform interventions are frequently and significantly influenced by development partnersrsquo priorities (Water Aid 2011) especially where transfers are collectively greater than domestic sources of financing generated through taxes and tariffs and where transfers are made outside the national policy agenda

As Figure 1 shows lsquodonor aid to the WASH sector as a percentage of GDP is higher than government budget allocations for WASH in Cambodia Ghana Liberia Madagascar Rwanda Timor-Leste and Uganda indicating both a donor-dominated sector and also that significant amounts of aid to the WASH sector in these countries is not recorded in central government budgets and accounts or is off-budgetrsquo (WaterAid 2011 p35) This disparity in international and domestic funding translates into disproportionate levels of influence by development partners in shaping national and sub-national development agenda priorities (WaterAid 2011)

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 3: Change in complex adaptive systems

Deirdre Casella Simone van Tongeren Igor Nikolic

Change in complex adaptive systems A review of concepts theory and approaches for tackling lsquowickedrsquo problems in achieving sustainable rural water services

Contents

ABBREVIATIONS 4

1 PURPOSE OF THIS REVIEW 5

2 WATER SERVICES A SYSTEM DESCRIPTION 721 Rural water services ndash from governance to systems perspectives 722 International development aid Exogenous change pressure 923 Capital investments but failing services 1224 Calls to re-think international development aid 12

3 CONCEPTS FROM THE COMPLEXITY SCIENCES 1431 Complex adaptive systems and the socio-technical system 1432 Change in socio-technical systems 1533 Universal Darwinism A meta-theory of evolving systems 1734 The language of institutions IAD and ADICO 1835 Rural water services as a socio-technical system 19

4 APPROACHES TO WHOLE SYSTEM CHANGE 2241 Collective impact Creating large-scale social change 2242 Learning alliances 2343 Problem-driven iterative adaptation 25

5 SIMULATING COMPLEX PROBLEMS TOOLS TO TEST AND LEARN 26 51 AGENT-BASED MODELLING 2652 SERIOUS GAMES 28

6 DISCUSSION AND CONCLUSION 30

REFERENCES 32

TablesTable 1 Agent-based model conceptualisation Rules of local government 29

FiguresFigure 1 Development partner funding and national allocations to WASH as percentage of GDP 10Figure 2 Economics of institutions 11Figure 3 IAD framework 19Figure 4 Republic of Uganda rural water sector actor landscape 21Figure 5 Sector change and innovation process 23Figure 6 Learning alliance approach to scaling change across institutional levels 24Figure 7 Structure of an agent-based model 27

4

Abbreviations

ADIC ADICO AIC CAS CBO GDP IAD NGO OECD PDIA SOHO WASH

attributes deontic aim conditionattributes deontic aim condition or elseattributes aim conditioncomplex adaptive systemcommunity-based organisationgross domestic productinstitutional analysis and developmentnon-governmental organisationOrganisation for Economic Co-operation and Developmentproblem-driven iterative adaptationself-organising hierarchical openwater sanitation and hygiene

5

1 Purpose of this review

This working paper a product of the Triple-S Water Services That Last project 1 is part of an effort to ground IRCrsquos approach to driving and supporting change processes in the water sanitation and hygiene (WASH) sector in the theory and methods of complexity sciences and systems thinking The review of literature from these fields identifies theory rationale and methods underpinning complexity-informed approaches to effecting change in large dynamic complex adaptive systems 2 The insights gained from this body of literature are discussed in relation to the rural water sector in low- and middle-income countriesmdasha sector facing lsquowickedrsquo 3 problems whose solution requires changes in the mindset and behaviours of multiple actors (Rittel and Webber 1973)

The domain of inquiry of this review is the rural water sector in low- and middle-income countries Section 2 describes a series of generally recognisable attributes of the rural water sector as they feature in low- and middle-income countries Salient attributes include the roles and responsibilities of key actors and patterns of interaction between and among key actors institutions and technology in this domain This section also reflects on the multi-faceted and inextricably linked challenges to delivering sustainable services and on the role of international development aid as an external influence in national development agendas and thereby in how public services are financed planned and implemented It reviews the literature on how international development aid can address these challenges by embracing concepts and practices informed by systems thinking and complexity sciences

Section 3 presents concepts and approaches from the complexity sciences and related fields including socio-technical systems as a specific type of complex adaptive system institutional analysis and whole-system change and explores whether they offer new perspectives on alternatives to prevailing national development and water service delivery policy and practice in particular Following the reflection on insights offered by the complexity sciences Section 3 concludes by reconsidering the rural water sector in light of the theory about complex adaptive systems and how change arises in such systems

Section 4 discusses three approaches featured in the literature for their potential to effect systemic change in complex adaptive systems as proposed alternatives to current policy making and planning in international development Given that large-scale systemic change can take several decades Section 5 presents tools for simulating change processes in complex adaptive systems for the purpose of envisaging exploring and experimenting with policy and implementation alternatives Section 6 concludes with a discussion of the insights gained through this review and identifies gaps in the literature about how to foster national systems that can deliver sustainable public services

1 Triple-S Water Services That Last is a six-year multi-country learning initiative to improve water supply to the rural poor that is led by IRC a Netherlands-based mission driven lsquothink and do tankrsquo pursuing the vision of WASH services for everyone forever The initiative is funded by the Bill amp Melinda Gates Foundation and has country programmes active in Ghana and Uganda with smaller initiatives supported in Burkina Faso Mozambique India and Honduras Lessons learned from work in countries feeds up to the international level where Triple-S promotes a re-appraisal of how development assistance to the rural water supply sector is designed and implemented

2 Moriarty Lockwood Carriger and Duti series of four blogposts httpwaterservicesthatlastwordpresscom20140224changing-the-whole-system-to-provide-water-sanitation-and-hygiene-services-that-last MarchndashMay 2014

3 lsquoWickedrsquo describes intractable not-easily-solved problemsmdashsuch as climate change pandemics poverty or natural disastersmdashthat require changes in mindset and behaviours of a great number of actors (Rittel and Webber 1973)

6

WORKING PAPER Change in complex adaptive systems

7

December 2015

In this section the rural water sub-sector and the challenges it faces in providing sustainable services are described Section 21 looks at the administrative and operational arrangements as well as roles and functions of actors involved in rural water sector in low- and middle-income countries Section 22 presents literature about the role and influence of international development as an external change pressure that influences domestic policy financial and technical aspects of rural water services in low- and middle-income countries Section 23 provides an overview of the multiple interlinked and intractable lsquowickedrsquo challenges to delivering services reflecting on the fact that the solutions are not hardware based Section 24 reviews literature arguing that such wicked challenges require alternative approaches grounded in the complexity sciences

21 RURAL WATER SERVICES ndash FROM GOVERNANCE TO SYSTEMS PERSPECTIVES

The overarching purpose of a national WASH sector is to develop and deliver sustainable water sanitation and hygiene services to users for domestic and productive purposes (Plummer and Slaymaker 2007) Although the WASH sector is not clearly demarcated as an entity with precisely identifiable boundaries a number of salient attributes can be identified across different national and regional contexts that enable us to consider the sector a lsquosystemrsquo There are also a number of identifiable activities that people must undertake to ensure that water services are delivered These attributes activities and interactions taken together give rise to a recognisable pattern that when effectively functioning results in the delivery of water services In this section the rural water sector and its salient attributes are described as the domain of focus of this literature review

In terms of attributes there are identifiable legal governance and organisational arrangements in place within nation states that guide the delivery of water services (Rogers and Hall 2003) In general a line ministry or department is the highest mandated authority responsible for ensuring that these services are delivered to all citizens In addition to developing and managing the natural water resource this national authority is tasked with developing and delivering water services to the population for domestic and productive purposes While the national authority for water resources management and

service delivery may also hold the remit for development and delivery of sanitation and hygiene services for the purpose of this review we focus in particular on the legal and administrative arrangements pertaining to rural water service delivery

In the context of decentralised water service delivery models in addition to the national government many other organisations and actors interacting on a continual basis are involved in policy making financial planning and management regulation and service provision activities across multiple administrative levels (Rogers and Hall 2003) Depending upon the national context these actors include sub-national government entities (eg provinces regions districts communes zones municipalities woredas panchayats) water utilities non-governmental organisations (NGOs) users community representatives private operators and capacity-building and financing bodies (de la Harpe 2007) Among this host of actors roles and responsibilities can be differentiated As highlighted by Smits et al (2011) a key distinction is the role of the service authority versus that of the service provider

The service authority generally a government body holds the legal responsibility for service delivery planning coordination regulation and oversight activities as well as technical assistance to water service providers (Smits et al 2011 Lockwood and Smits 2011) In contrast the service provider is the organisation or individual responsible for day-to-day water service which includes the operation maintenance and administration of the water system

How the water service provision role is fulfilled varies widely Most countries have a range of service provision options or service delivery models whereby lsquothe service authority can opt to provide services itself (through a municipal department or municipal company) or hellip delegate this responsibility by contracting an outside agency such as a community-based organisation (CBO) private operator public sector utility or company or non-governmental organisation (NGO) who in turn may hire a private person (plumber or mechanic) to carry out parts of the workrsquo (Smits et al 2011 p5)

In addition to the legal administrative and organisational attributes and interactionsmdashthat is the lsquosocialrsquo components of a water servicemdashthere is also

2 Water services A system description

8

WORKING PAPER Change in complex adaptive systems

the technical component The physical infrastructure required for the delivery of water supply varies greatly among and even within service areas4 depending upon factors such as geographical and hydrological conditions preferences for certain technologies available financial resources and population size and density

Whether the infrastructure is a stand-alone hand pump or a networked gravity-fed piped scheme formal and informal arrangements among the authority provider users civil society and international development organisations are required to ensure sustainable water services (Keohane and Ostrom 1995 Rogers and Hall 2003 Plummer and Slaymaker 2007) These arrangements entail policy- and decision-making processes about responsibilities and actor relationships through which the power responsibilities norms values and formal agreements embedded in laws and policies are negotiated among and implemented by the array of stakeholders whose roles and responsibilities may overlap (Ostrom and Janssen 2004)

This view of interdependent networks of multiple actors or agents interacting across multiple administrative levels embodies a lsquogovernance perspectiversquo (Kooiman 1993 Stoker 1998 DFID 2007) In reflecting on how to provide for a collective interest such as the provision of public services the governance perspective offers a lsquoframework for understanding changing processes of governing characterised by processes of adaptation learning and experimentrsquo (Stoker 1998 p18) This perspective helps to conceptualise how functions related to service delivery are dispersed over a wide array of actors organisations and coordination platforms spanning different national development sectors and administrative levels Notably from a governance perspective while (central) governments continue to play a role in how public services are provided as Bache (2003) notes this role is increasingly one of coordination and steering and is concomitant with an increase in the involvement of non-government actors in policy-making and service delivery The Global Water Partnership (GWP) defines water

governance as lsquothe range of political social economic and administrative systems that are in place to develop and manage water resources and the delivery of water services at different levels of societyrsquo (Rogers and Hall 2003) The literature on governance also highlights the importance of public participation in governance processes for the potential to lsquoimprove the quality of decision making by opening up the decision-making process and making better use of the information and creativity that is available in society improve public under-standing of the management issues at stake make decision making more transparent and might stimulate the different government bodies involved to coordinate their actions more in order to provide serious follow-up to the inputs receivedrsquo and potentially strengthen democratic processes where government does not have all the resources required to lsquomanage an issue effectivelyrsquo (Huitema et al 2009 p5)

The prevailing governance approaches in low- and middle-income countries are context specific and have evolved over long periods of time in response to change pressures such as political social and economic processes (Plummer and Slaymaker 2007) Irrespective of the governance arrangements in a given national context the governance perspective as described here makes it possible to introduce a systems perspective5 to understanding the rural water sector6 This line of inquiry will be addressed in further detail in Section 3

By framing water services as a system that is open to feedback from its environment it becomes possible to consider external change pressures increasingly recognised for their effect on how national development agendasmdashincluding water service deliverymdashare formed and implemented One such pressure is the policy and finance support provided to low- and middle-income countries in the form of international development aid from parties such as UN agencies international finance institutes philanthropic organisations non-governmental organisations and middle- and high-income countries (Rogers and Hall 2003 Mowles et al 2008

4 A service area is the area of jurisdiction and population covered by a service authority Service areas are typically linked to the boundaries of human settlement (towns villages hamlets and scattered rural settlements) but may not correspond precisely with administrative boundaries (IRC Glossary accessed 16 October 2014)

5 A systems perspective takes into account all of the behaviours of a system as a whole in the context of its environment and is lsquoa non-reductionist approach to describing the properties of the system itselfrsquo A description of the whole must include an explanation of the relationships between the parts as well as any additional information needed to describe the behaviour of the entire system (after Bar-Yam 1997 2005)

6 A system is defined by Ryan (2008) as lsquoa representation of an entity as a complex whole open to feedback from its environmentrsquo Ryan (2008) Burke (2006) and other authors on systems and complexity sciences make the important observation that such representations are idealisations based on simplified assumptions Thus although they offer a valid means for identifying and analysing an entity and its dynamics lsquothere are limits to their applicationrsquo (Mowles et al 2008 Ryan 2008)

9

December 2015

Ramalingam 2013) Given the importance of international development aid in shaping national development agenda priorities the next section summarises literature discussing its role and reflects on the dynamic arising in the water sector where it is a main source of finance for service development and a highly influential exogenous change pressure

22 INTERNATIONAL DEVELOPMENT AID EXOGENOUS CHANGE PRESSURE

In this sub-section the role of international development aid in shaping national development agendas of low- and middle-income countries is discussed The international development aid system provides policy financial and technical support to developing nations for education health transportation energy and local and regional economic development and trade A clear understanding of the dynamic created by the involvement of such influential external agents is critical to understanding the current challenges to achieving sustainable water services let alone the realisation of resilient national systems that can develop and deliver the public services required for sustainable and equitable social and economic development (Mowles et al 2008) Many researchers have explored how the current architecture of international development aid delivery is hindering the potential of nations to achieve these social and economic development goals

When developing countries need capital to build infrastructure for public services international development aid actorsmdashranging from development banks funders and bi-lateral government agencies to NGOs and philanthropic organisations (henceforth lsquodevelopment partnersrsquo)mdashmay explicitly partner with a recipient government through policy budget andor technical support strategies and implementation plans (Mowles et al 2008 Ramalingam 2013) It is not uncommon for development partners to bypass nationally led processes and directly implement programmes at the user and community levels (Nimanya et al 2011)

The visions missions and mandates of development partners vary greatly and determine the nature of their interactions with domestic partners including financial investment decisions In addition to investing in infrastructure development partners may support organisational policy and capacity aspects related to the sustainable delivery of a service including both direct support (monitoring maintenance repairs replacements training of staff)

and indirect support (macro-level planning and policy making) These essential components of sustainable water services known as post-construction support or lsquosoftwarersquo are as important as the infrastructure lsquohardwarersquo yet often neglected with actual levels of financial support considered insufficient (Rogers and Hall 2003 Smits et al 2011)

Reasons for this neglect include the desire to focus resources on increasing coverage rates for unserved populations (WHO 2012) perceptions about the risk of corruption faulty assumptions about the lsquobestrsquo governance arrangements for post-construction activities (Schouten and Moriarty 2003) and the desire to see tangible easily measurable results from an investment (Garandeau et al 2009)

Understanding the sources of financing provides insight into how WASH policy priorities are determined The Organisation for Economic Co-operation and Development (OECD 2009) distinguishes the sources of financing for WASH services as the three Ts tariffs which are contributions paid by service users transfers in the form of assistance from development partners and taxes which are levied by national or regional governments In many countries the funding from international aid (transfers to use the OECD nomenclature) is at least as much as the funding from the two domestic sources tariffs and taxes (Figure 1)

Financing from transfers is not problematic on its own and in fact does tremendous good in many countries However since lsquohe who pays the piper calls the tunersquo national policy strategy and governance reform interventions are frequently and significantly influenced by development partnersrsquo priorities (Water Aid 2011) especially where transfers are collectively greater than domestic sources of financing generated through taxes and tariffs and where transfers are made outside the national policy agenda

As Figure 1 shows lsquodonor aid to the WASH sector as a percentage of GDP is higher than government budget allocations for WASH in Cambodia Ghana Liberia Madagascar Rwanda Timor-Leste and Uganda indicating both a donor-dominated sector and also that significant amounts of aid to the WASH sector in these countries is not recorded in central government budgets and accounts or is off-budgetrsquo (WaterAid 2011 p35) This disparity in international and domestic funding translates into disproportionate levels of influence by development partners in shaping national and sub-national development agenda priorities (WaterAid 2011)

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 4: Change in complex adaptive systems

Contents

ABBREVIATIONS 4

1 PURPOSE OF THIS REVIEW 5

2 WATER SERVICES A SYSTEM DESCRIPTION 721 Rural water services ndash from governance to systems perspectives 722 International development aid Exogenous change pressure 923 Capital investments but failing services 1224 Calls to re-think international development aid 12

3 CONCEPTS FROM THE COMPLEXITY SCIENCES 1431 Complex adaptive systems and the socio-technical system 1432 Change in socio-technical systems 1533 Universal Darwinism A meta-theory of evolving systems 1734 The language of institutions IAD and ADICO 1835 Rural water services as a socio-technical system 19

4 APPROACHES TO WHOLE SYSTEM CHANGE 2241 Collective impact Creating large-scale social change 2242 Learning alliances 2343 Problem-driven iterative adaptation 25

5 SIMULATING COMPLEX PROBLEMS TOOLS TO TEST AND LEARN 26 51 AGENT-BASED MODELLING 2652 SERIOUS GAMES 28

6 DISCUSSION AND CONCLUSION 30

REFERENCES 32

TablesTable 1 Agent-based model conceptualisation Rules of local government 29

FiguresFigure 1 Development partner funding and national allocations to WASH as percentage of GDP 10Figure 2 Economics of institutions 11Figure 3 IAD framework 19Figure 4 Republic of Uganda rural water sector actor landscape 21Figure 5 Sector change and innovation process 23Figure 6 Learning alliance approach to scaling change across institutional levels 24Figure 7 Structure of an agent-based model 27

4

Abbreviations

ADIC ADICO AIC CAS CBO GDP IAD NGO OECD PDIA SOHO WASH

attributes deontic aim conditionattributes deontic aim condition or elseattributes aim conditioncomplex adaptive systemcommunity-based organisationgross domestic productinstitutional analysis and developmentnon-governmental organisationOrganisation for Economic Co-operation and Developmentproblem-driven iterative adaptationself-organising hierarchical openwater sanitation and hygiene

5

1 Purpose of this review

This working paper a product of the Triple-S Water Services That Last project 1 is part of an effort to ground IRCrsquos approach to driving and supporting change processes in the water sanitation and hygiene (WASH) sector in the theory and methods of complexity sciences and systems thinking The review of literature from these fields identifies theory rationale and methods underpinning complexity-informed approaches to effecting change in large dynamic complex adaptive systems 2 The insights gained from this body of literature are discussed in relation to the rural water sector in low- and middle-income countriesmdasha sector facing lsquowickedrsquo 3 problems whose solution requires changes in the mindset and behaviours of multiple actors (Rittel and Webber 1973)

The domain of inquiry of this review is the rural water sector in low- and middle-income countries Section 2 describes a series of generally recognisable attributes of the rural water sector as they feature in low- and middle-income countries Salient attributes include the roles and responsibilities of key actors and patterns of interaction between and among key actors institutions and technology in this domain This section also reflects on the multi-faceted and inextricably linked challenges to delivering sustainable services and on the role of international development aid as an external influence in national development agendas and thereby in how public services are financed planned and implemented It reviews the literature on how international development aid can address these challenges by embracing concepts and practices informed by systems thinking and complexity sciences

Section 3 presents concepts and approaches from the complexity sciences and related fields including socio-technical systems as a specific type of complex adaptive system institutional analysis and whole-system change and explores whether they offer new perspectives on alternatives to prevailing national development and water service delivery policy and practice in particular Following the reflection on insights offered by the complexity sciences Section 3 concludes by reconsidering the rural water sector in light of the theory about complex adaptive systems and how change arises in such systems

Section 4 discusses three approaches featured in the literature for their potential to effect systemic change in complex adaptive systems as proposed alternatives to current policy making and planning in international development Given that large-scale systemic change can take several decades Section 5 presents tools for simulating change processes in complex adaptive systems for the purpose of envisaging exploring and experimenting with policy and implementation alternatives Section 6 concludes with a discussion of the insights gained through this review and identifies gaps in the literature about how to foster national systems that can deliver sustainable public services

1 Triple-S Water Services That Last is a six-year multi-country learning initiative to improve water supply to the rural poor that is led by IRC a Netherlands-based mission driven lsquothink and do tankrsquo pursuing the vision of WASH services for everyone forever The initiative is funded by the Bill amp Melinda Gates Foundation and has country programmes active in Ghana and Uganda with smaller initiatives supported in Burkina Faso Mozambique India and Honduras Lessons learned from work in countries feeds up to the international level where Triple-S promotes a re-appraisal of how development assistance to the rural water supply sector is designed and implemented

2 Moriarty Lockwood Carriger and Duti series of four blogposts httpwaterservicesthatlastwordpresscom20140224changing-the-whole-system-to-provide-water-sanitation-and-hygiene-services-that-last MarchndashMay 2014

3 lsquoWickedrsquo describes intractable not-easily-solved problemsmdashsuch as climate change pandemics poverty or natural disastersmdashthat require changes in mindset and behaviours of a great number of actors (Rittel and Webber 1973)

6

WORKING PAPER Change in complex adaptive systems

7

December 2015

In this section the rural water sub-sector and the challenges it faces in providing sustainable services are described Section 21 looks at the administrative and operational arrangements as well as roles and functions of actors involved in rural water sector in low- and middle-income countries Section 22 presents literature about the role and influence of international development as an external change pressure that influences domestic policy financial and technical aspects of rural water services in low- and middle-income countries Section 23 provides an overview of the multiple interlinked and intractable lsquowickedrsquo challenges to delivering services reflecting on the fact that the solutions are not hardware based Section 24 reviews literature arguing that such wicked challenges require alternative approaches grounded in the complexity sciences

21 RURAL WATER SERVICES ndash FROM GOVERNANCE TO SYSTEMS PERSPECTIVES

The overarching purpose of a national WASH sector is to develop and deliver sustainable water sanitation and hygiene services to users for domestic and productive purposes (Plummer and Slaymaker 2007) Although the WASH sector is not clearly demarcated as an entity with precisely identifiable boundaries a number of salient attributes can be identified across different national and regional contexts that enable us to consider the sector a lsquosystemrsquo There are also a number of identifiable activities that people must undertake to ensure that water services are delivered These attributes activities and interactions taken together give rise to a recognisable pattern that when effectively functioning results in the delivery of water services In this section the rural water sector and its salient attributes are described as the domain of focus of this literature review

In terms of attributes there are identifiable legal governance and organisational arrangements in place within nation states that guide the delivery of water services (Rogers and Hall 2003) In general a line ministry or department is the highest mandated authority responsible for ensuring that these services are delivered to all citizens In addition to developing and managing the natural water resource this national authority is tasked with developing and delivering water services to the population for domestic and productive purposes While the national authority for water resources management and

service delivery may also hold the remit for development and delivery of sanitation and hygiene services for the purpose of this review we focus in particular on the legal and administrative arrangements pertaining to rural water service delivery

In the context of decentralised water service delivery models in addition to the national government many other organisations and actors interacting on a continual basis are involved in policy making financial planning and management regulation and service provision activities across multiple administrative levels (Rogers and Hall 2003) Depending upon the national context these actors include sub-national government entities (eg provinces regions districts communes zones municipalities woredas panchayats) water utilities non-governmental organisations (NGOs) users community representatives private operators and capacity-building and financing bodies (de la Harpe 2007) Among this host of actors roles and responsibilities can be differentiated As highlighted by Smits et al (2011) a key distinction is the role of the service authority versus that of the service provider

The service authority generally a government body holds the legal responsibility for service delivery planning coordination regulation and oversight activities as well as technical assistance to water service providers (Smits et al 2011 Lockwood and Smits 2011) In contrast the service provider is the organisation or individual responsible for day-to-day water service which includes the operation maintenance and administration of the water system

How the water service provision role is fulfilled varies widely Most countries have a range of service provision options or service delivery models whereby lsquothe service authority can opt to provide services itself (through a municipal department or municipal company) or hellip delegate this responsibility by contracting an outside agency such as a community-based organisation (CBO) private operator public sector utility or company or non-governmental organisation (NGO) who in turn may hire a private person (plumber or mechanic) to carry out parts of the workrsquo (Smits et al 2011 p5)

In addition to the legal administrative and organisational attributes and interactionsmdashthat is the lsquosocialrsquo components of a water servicemdashthere is also

2 Water services A system description

8

WORKING PAPER Change in complex adaptive systems

the technical component The physical infrastructure required for the delivery of water supply varies greatly among and even within service areas4 depending upon factors such as geographical and hydrological conditions preferences for certain technologies available financial resources and population size and density

Whether the infrastructure is a stand-alone hand pump or a networked gravity-fed piped scheme formal and informal arrangements among the authority provider users civil society and international development organisations are required to ensure sustainable water services (Keohane and Ostrom 1995 Rogers and Hall 2003 Plummer and Slaymaker 2007) These arrangements entail policy- and decision-making processes about responsibilities and actor relationships through which the power responsibilities norms values and formal agreements embedded in laws and policies are negotiated among and implemented by the array of stakeholders whose roles and responsibilities may overlap (Ostrom and Janssen 2004)

This view of interdependent networks of multiple actors or agents interacting across multiple administrative levels embodies a lsquogovernance perspectiversquo (Kooiman 1993 Stoker 1998 DFID 2007) In reflecting on how to provide for a collective interest such as the provision of public services the governance perspective offers a lsquoframework for understanding changing processes of governing characterised by processes of adaptation learning and experimentrsquo (Stoker 1998 p18) This perspective helps to conceptualise how functions related to service delivery are dispersed over a wide array of actors organisations and coordination platforms spanning different national development sectors and administrative levels Notably from a governance perspective while (central) governments continue to play a role in how public services are provided as Bache (2003) notes this role is increasingly one of coordination and steering and is concomitant with an increase in the involvement of non-government actors in policy-making and service delivery The Global Water Partnership (GWP) defines water

governance as lsquothe range of political social economic and administrative systems that are in place to develop and manage water resources and the delivery of water services at different levels of societyrsquo (Rogers and Hall 2003) The literature on governance also highlights the importance of public participation in governance processes for the potential to lsquoimprove the quality of decision making by opening up the decision-making process and making better use of the information and creativity that is available in society improve public under-standing of the management issues at stake make decision making more transparent and might stimulate the different government bodies involved to coordinate their actions more in order to provide serious follow-up to the inputs receivedrsquo and potentially strengthen democratic processes where government does not have all the resources required to lsquomanage an issue effectivelyrsquo (Huitema et al 2009 p5)

The prevailing governance approaches in low- and middle-income countries are context specific and have evolved over long periods of time in response to change pressures such as political social and economic processes (Plummer and Slaymaker 2007) Irrespective of the governance arrangements in a given national context the governance perspective as described here makes it possible to introduce a systems perspective5 to understanding the rural water sector6 This line of inquiry will be addressed in further detail in Section 3

By framing water services as a system that is open to feedback from its environment it becomes possible to consider external change pressures increasingly recognised for their effect on how national development agendasmdashincluding water service deliverymdashare formed and implemented One such pressure is the policy and finance support provided to low- and middle-income countries in the form of international development aid from parties such as UN agencies international finance institutes philanthropic organisations non-governmental organisations and middle- and high-income countries (Rogers and Hall 2003 Mowles et al 2008

4 A service area is the area of jurisdiction and population covered by a service authority Service areas are typically linked to the boundaries of human settlement (towns villages hamlets and scattered rural settlements) but may not correspond precisely with administrative boundaries (IRC Glossary accessed 16 October 2014)

5 A systems perspective takes into account all of the behaviours of a system as a whole in the context of its environment and is lsquoa non-reductionist approach to describing the properties of the system itselfrsquo A description of the whole must include an explanation of the relationships between the parts as well as any additional information needed to describe the behaviour of the entire system (after Bar-Yam 1997 2005)

6 A system is defined by Ryan (2008) as lsquoa representation of an entity as a complex whole open to feedback from its environmentrsquo Ryan (2008) Burke (2006) and other authors on systems and complexity sciences make the important observation that such representations are idealisations based on simplified assumptions Thus although they offer a valid means for identifying and analysing an entity and its dynamics lsquothere are limits to their applicationrsquo (Mowles et al 2008 Ryan 2008)

9

December 2015

Ramalingam 2013) Given the importance of international development aid in shaping national development agenda priorities the next section summarises literature discussing its role and reflects on the dynamic arising in the water sector where it is a main source of finance for service development and a highly influential exogenous change pressure

22 INTERNATIONAL DEVELOPMENT AID EXOGENOUS CHANGE PRESSURE

In this sub-section the role of international development aid in shaping national development agendas of low- and middle-income countries is discussed The international development aid system provides policy financial and technical support to developing nations for education health transportation energy and local and regional economic development and trade A clear understanding of the dynamic created by the involvement of such influential external agents is critical to understanding the current challenges to achieving sustainable water services let alone the realisation of resilient national systems that can develop and deliver the public services required for sustainable and equitable social and economic development (Mowles et al 2008) Many researchers have explored how the current architecture of international development aid delivery is hindering the potential of nations to achieve these social and economic development goals

When developing countries need capital to build infrastructure for public services international development aid actorsmdashranging from development banks funders and bi-lateral government agencies to NGOs and philanthropic organisations (henceforth lsquodevelopment partnersrsquo)mdashmay explicitly partner with a recipient government through policy budget andor technical support strategies and implementation plans (Mowles et al 2008 Ramalingam 2013) It is not uncommon for development partners to bypass nationally led processes and directly implement programmes at the user and community levels (Nimanya et al 2011)

The visions missions and mandates of development partners vary greatly and determine the nature of their interactions with domestic partners including financial investment decisions In addition to investing in infrastructure development partners may support organisational policy and capacity aspects related to the sustainable delivery of a service including both direct support (monitoring maintenance repairs replacements training of staff)

and indirect support (macro-level planning and policy making) These essential components of sustainable water services known as post-construction support or lsquosoftwarersquo are as important as the infrastructure lsquohardwarersquo yet often neglected with actual levels of financial support considered insufficient (Rogers and Hall 2003 Smits et al 2011)

Reasons for this neglect include the desire to focus resources on increasing coverage rates for unserved populations (WHO 2012) perceptions about the risk of corruption faulty assumptions about the lsquobestrsquo governance arrangements for post-construction activities (Schouten and Moriarty 2003) and the desire to see tangible easily measurable results from an investment (Garandeau et al 2009)

Understanding the sources of financing provides insight into how WASH policy priorities are determined The Organisation for Economic Co-operation and Development (OECD 2009) distinguishes the sources of financing for WASH services as the three Ts tariffs which are contributions paid by service users transfers in the form of assistance from development partners and taxes which are levied by national or regional governments In many countries the funding from international aid (transfers to use the OECD nomenclature) is at least as much as the funding from the two domestic sources tariffs and taxes (Figure 1)

Financing from transfers is not problematic on its own and in fact does tremendous good in many countries However since lsquohe who pays the piper calls the tunersquo national policy strategy and governance reform interventions are frequently and significantly influenced by development partnersrsquo priorities (Water Aid 2011) especially where transfers are collectively greater than domestic sources of financing generated through taxes and tariffs and where transfers are made outside the national policy agenda

As Figure 1 shows lsquodonor aid to the WASH sector as a percentage of GDP is higher than government budget allocations for WASH in Cambodia Ghana Liberia Madagascar Rwanda Timor-Leste and Uganda indicating both a donor-dominated sector and also that significant amounts of aid to the WASH sector in these countries is not recorded in central government budgets and accounts or is off-budgetrsquo (WaterAid 2011 p35) This disparity in international and domestic funding translates into disproportionate levels of influence by development partners in shaping national and sub-national development agenda priorities (WaterAid 2011)

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 5: Change in complex adaptive systems

4

Abbreviations

ADIC ADICO AIC CAS CBO GDP IAD NGO OECD PDIA SOHO WASH

attributes deontic aim conditionattributes deontic aim condition or elseattributes aim conditioncomplex adaptive systemcommunity-based organisationgross domestic productinstitutional analysis and developmentnon-governmental organisationOrganisation for Economic Co-operation and Developmentproblem-driven iterative adaptationself-organising hierarchical openwater sanitation and hygiene

5

1 Purpose of this review

This working paper a product of the Triple-S Water Services That Last project 1 is part of an effort to ground IRCrsquos approach to driving and supporting change processes in the water sanitation and hygiene (WASH) sector in the theory and methods of complexity sciences and systems thinking The review of literature from these fields identifies theory rationale and methods underpinning complexity-informed approaches to effecting change in large dynamic complex adaptive systems 2 The insights gained from this body of literature are discussed in relation to the rural water sector in low- and middle-income countriesmdasha sector facing lsquowickedrsquo 3 problems whose solution requires changes in the mindset and behaviours of multiple actors (Rittel and Webber 1973)

The domain of inquiry of this review is the rural water sector in low- and middle-income countries Section 2 describes a series of generally recognisable attributes of the rural water sector as they feature in low- and middle-income countries Salient attributes include the roles and responsibilities of key actors and patterns of interaction between and among key actors institutions and technology in this domain This section also reflects on the multi-faceted and inextricably linked challenges to delivering sustainable services and on the role of international development aid as an external influence in national development agendas and thereby in how public services are financed planned and implemented It reviews the literature on how international development aid can address these challenges by embracing concepts and practices informed by systems thinking and complexity sciences

Section 3 presents concepts and approaches from the complexity sciences and related fields including socio-technical systems as a specific type of complex adaptive system institutional analysis and whole-system change and explores whether they offer new perspectives on alternatives to prevailing national development and water service delivery policy and practice in particular Following the reflection on insights offered by the complexity sciences Section 3 concludes by reconsidering the rural water sector in light of the theory about complex adaptive systems and how change arises in such systems

Section 4 discusses three approaches featured in the literature for their potential to effect systemic change in complex adaptive systems as proposed alternatives to current policy making and planning in international development Given that large-scale systemic change can take several decades Section 5 presents tools for simulating change processes in complex adaptive systems for the purpose of envisaging exploring and experimenting with policy and implementation alternatives Section 6 concludes with a discussion of the insights gained through this review and identifies gaps in the literature about how to foster national systems that can deliver sustainable public services

1 Triple-S Water Services That Last is a six-year multi-country learning initiative to improve water supply to the rural poor that is led by IRC a Netherlands-based mission driven lsquothink and do tankrsquo pursuing the vision of WASH services for everyone forever The initiative is funded by the Bill amp Melinda Gates Foundation and has country programmes active in Ghana and Uganda with smaller initiatives supported in Burkina Faso Mozambique India and Honduras Lessons learned from work in countries feeds up to the international level where Triple-S promotes a re-appraisal of how development assistance to the rural water supply sector is designed and implemented

2 Moriarty Lockwood Carriger and Duti series of four blogposts httpwaterservicesthatlastwordpresscom20140224changing-the-whole-system-to-provide-water-sanitation-and-hygiene-services-that-last MarchndashMay 2014

3 lsquoWickedrsquo describes intractable not-easily-solved problemsmdashsuch as climate change pandemics poverty or natural disastersmdashthat require changes in mindset and behaviours of a great number of actors (Rittel and Webber 1973)

6

WORKING PAPER Change in complex adaptive systems

7

December 2015

In this section the rural water sub-sector and the challenges it faces in providing sustainable services are described Section 21 looks at the administrative and operational arrangements as well as roles and functions of actors involved in rural water sector in low- and middle-income countries Section 22 presents literature about the role and influence of international development as an external change pressure that influences domestic policy financial and technical aspects of rural water services in low- and middle-income countries Section 23 provides an overview of the multiple interlinked and intractable lsquowickedrsquo challenges to delivering services reflecting on the fact that the solutions are not hardware based Section 24 reviews literature arguing that such wicked challenges require alternative approaches grounded in the complexity sciences

21 RURAL WATER SERVICES ndash FROM GOVERNANCE TO SYSTEMS PERSPECTIVES

The overarching purpose of a national WASH sector is to develop and deliver sustainable water sanitation and hygiene services to users for domestic and productive purposes (Plummer and Slaymaker 2007) Although the WASH sector is not clearly demarcated as an entity with precisely identifiable boundaries a number of salient attributes can be identified across different national and regional contexts that enable us to consider the sector a lsquosystemrsquo There are also a number of identifiable activities that people must undertake to ensure that water services are delivered These attributes activities and interactions taken together give rise to a recognisable pattern that when effectively functioning results in the delivery of water services In this section the rural water sector and its salient attributes are described as the domain of focus of this literature review

In terms of attributes there are identifiable legal governance and organisational arrangements in place within nation states that guide the delivery of water services (Rogers and Hall 2003) In general a line ministry or department is the highest mandated authority responsible for ensuring that these services are delivered to all citizens In addition to developing and managing the natural water resource this national authority is tasked with developing and delivering water services to the population for domestic and productive purposes While the national authority for water resources management and

service delivery may also hold the remit for development and delivery of sanitation and hygiene services for the purpose of this review we focus in particular on the legal and administrative arrangements pertaining to rural water service delivery

In the context of decentralised water service delivery models in addition to the national government many other organisations and actors interacting on a continual basis are involved in policy making financial planning and management regulation and service provision activities across multiple administrative levels (Rogers and Hall 2003) Depending upon the national context these actors include sub-national government entities (eg provinces regions districts communes zones municipalities woredas panchayats) water utilities non-governmental organisations (NGOs) users community representatives private operators and capacity-building and financing bodies (de la Harpe 2007) Among this host of actors roles and responsibilities can be differentiated As highlighted by Smits et al (2011) a key distinction is the role of the service authority versus that of the service provider

The service authority generally a government body holds the legal responsibility for service delivery planning coordination regulation and oversight activities as well as technical assistance to water service providers (Smits et al 2011 Lockwood and Smits 2011) In contrast the service provider is the organisation or individual responsible for day-to-day water service which includes the operation maintenance and administration of the water system

How the water service provision role is fulfilled varies widely Most countries have a range of service provision options or service delivery models whereby lsquothe service authority can opt to provide services itself (through a municipal department or municipal company) or hellip delegate this responsibility by contracting an outside agency such as a community-based organisation (CBO) private operator public sector utility or company or non-governmental organisation (NGO) who in turn may hire a private person (plumber or mechanic) to carry out parts of the workrsquo (Smits et al 2011 p5)

In addition to the legal administrative and organisational attributes and interactionsmdashthat is the lsquosocialrsquo components of a water servicemdashthere is also

2 Water services A system description

8

WORKING PAPER Change in complex adaptive systems

the technical component The physical infrastructure required for the delivery of water supply varies greatly among and even within service areas4 depending upon factors such as geographical and hydrological conditions preferences for certain technologies available financial resources and population size and density

Whether the infrastructure is a stand-alone hand pump or a networked gravity-fed piped scheme formal and informal arrangements among the authority provider users civil society and international development organisations are required to ensure sustainable water services (Keohane and Ostrom 1995 Rogers and Hall 2003 Plummer and Slaymaker 2007) These arrangements entail policy- and decision-making processes about responsibilities and actor relationships through which the power responsibilities norms values and formal agreements embedded in laws and policies are negotiated among and implemented by the array of stakeholders whose roles and responsibilities may overlap (Ostrom and Janssen 2004)

This view of interdependent networks of multiple actors or agents interacting across multiple administrative levels embodies a lsquogovernance perspectiversquo (Kooiman 1993 Stoker 1998 DFID 2007) In reflecting on how to provide for a collective interest such as the provision of public services the governance perspective offers a lsquoframework for understanding changing processes of governing characterised by processes of adaptation learning and experimentrsquo (Stoker 1998 p18) This perspective helps to conceptualise how functions related to service delivery are dispersed over a wide array of actors organisations and coordination platforms spanning different national development sectors and administrative levels Notably from a governance perspective while (central) governments continue to play a role in how public services are provided as Bache (2003) notes this role is increasingly one of coordination and steering and is concomitant with an increase in the involvement of non-government actors in policy-making and service delivery The Global Water Partnership (GWP) defines water

governance as lsquothe range of political social economic and administrative systems that are in place to develop and manage water resources and the delivery of water services at different levels of societyrsquo (Rogers and Hall 2003) The literature on governance also highlights the importance of public participation in governance processes for the potential to lsquoimprove the quality of decision making by opening up the decision-making process and making better use of the information and creativity that is available in society improve public under-standing of the management issues at stake make decision making more transparent and might stimulate the different government bodies involved to coordinate their actions more in order to provide serious follow-up to the inputs receivedrsquo and potentially strengthen democratic processes where government does not have all the resources required to lsquomanage an issue effectivelyrsquo (Huitema et al 2009 p5)

The prevailing governance approaches in low- and middle-income countries are context specific and have evolved over long periods of time in response to change pressures such as political social and economic processes (Plummer and Slaymaker 2007) Irrespective of the governance arrangements in a given national context the governance perspective as described here makes it possible to introduce a systems perspective5 to understanding the rural water sector6 This line of inquiry will be addressed in further detail in Section 3

By framing water services as a system that is open to feedback from its environment it becomes possible to consider external change pressures increasingly recognised for their effect on how national development agendasmdashincluding water service deliverymdashare formed and implemented One such pressure is the policy and finance support provided to low- and middle-income countries in the form of international development aid from parties such as UN agencies international finance institutes philanthropic organisations non-governmental organisations and middle- and high-income countries (Rogers and Hall 2003 Mowles et al 2008

4 A service area is the area of jurisdiction and population covered by a service authority Service areas are typically linked to the boundaries of human settlement (towns villages hamlets and scattered rural settlements) but may not correspond precisely with administrative boundaries (IRC Glossary accessed 16 October 2014)

5 A systems perspective takes into account all of the behaviours of a system as a whole in the context of its environment and is lsquoa non-reductionist approach to describing the properties of the system itselfrsquo A description of the whole must include an explanation of the relationships between the parts as well as any additional information needed to describe the behaviour of the entire system (after Bar-Yam 1997 2005)

6 A system is defined by Ryan (2008) as lsquoa representation of an entity as a complex whole open to feedback from its environmentrsquo Ryan (2008) Burke (2006) and other authors on systems and complexity sciences make the important observation that such representations are idealisations based on simplified assumptions Thus although they offer a valid means for identifying and analysing an entity and its dynamics lsquothere are limits to their applicationrsquo (Mowles et al 2008 Ryan 2008)

9

December 2015

Ramalingam 2013) Given the importance of international development aid in shaping national development agenda priorities the next section summarises literature discussing its role and reflects on the dynamic arising in the water sector where it is a main source of finance for service development and a highly influential exogenous change pressure

22 INTERNATIONAL DEVELOPMENT AID EXOGENOUS CHANGE PRESSURE

In this sub-section the role of international development aid in shaping national development agendas of low- and middle-income countries is discussed The international development aid system provides policy financial and technical support to developing nations for education health transportation energy and local and regional economic development and trade A clear understanding of the dynamic created by the involvement of such influential external agents is critical to understanding the current challenges to achieving sustainable water services let alone the realisation of resilient national systems that can develop and deliver the public services required for sustainable and equitable social and economic development (Mowles et al 2008) Many researchers have explored how the current architecture of international development aid delivery is hindering the potential of nations to achieve these social and economic development goals

When developing countries need capital to build infrastructure for public services international development aid actorsmdashranging from development banks funders and bi-lateral government agencies to NGOs and philanthropic organisations (henceforth lsquodevelopment partnersrsquo)mdashmay explicitly partner with a recipient government through policy budget andor technical support strategies and implementation plans (Mowles et al 2008 Ramalingam 2013) It is not uncommon for development partners to bypass nationally led processes and directly implement programmes at the user and community levels (Nimanya et al 2011)

The visions missions and mandates of development partners vary greatly and determine the nature of their interactions with domestic partners including financial investment decisions In addition to investing in infrastructure development partners may support organisational policy and capacity aspects related to the sustainable delivery of a service including both direct support (monitoring maintenance repairs replacements training of staff)

and indirect support (macro-level planning and policy making) These essential components of sustainable water services known as post-construction support or lsquosoftwarersquo are as important as the infrastructure lsquohardwarersquo yet often neglected with actual levels of financial support considered insufficient (Rogers and Hall 2003 Smits et al 2011)

Reasons for this neglect include the desire to focus resources on increasing coverage rates for unserved populations (WHO 2012) perceptions about the risk of corruption faulty assumptions about the lsquobestrsquo governance arrangements for post-construction activities (Schouten and Moriarty 2003) and the desire to see tangible easily measurable results from an investment (Garandeau et al 2009)

Understanding the sources of financing provides insight into how WASH policy priorities are determined The Organisation for Economic Co-operation and Development (OECD 2009) distinguishes the sources of financing for WASH services as the three Ts tariffs which are contributions paid by service users transfers in the form of assistance from development partners and taxes which are levied by national or regional governments In many countries the funding from international aid (transfers to use the OECD nomenclature) is at least as much as the funding from the two domestic sources tariffs and taxes (Figure 1)

Financing from transfers is not problematic on its own and in fact does tremendous good in many countries However since lsquohe who pays the piper calls the tunersquo national policy strategy and governance reform interventions are frequently and significantly influenced by development partnersrsquo priorities (Water Aid 2011) especially where transfers are collectively greater than domestic sources of financing generated through taxes and tariffs and where transfers are made outside the national policy agenda

As Figure 1 shows lsquodonor aid to the WASH sector as a percentage of GDP is higher than government budget allocations for WASH in Cambodia Ghana Liberia Madagascar Rwanda Timor-Leste and Uganda indicating both a donor-dominated sector and also that significant amounts of aid to the WASH sector in these countries is not recorded in central government budgets and accounts or is off-budgetrsquo (WaterAid 2011 p35) This disparity in international and domestic funding translates into disproportionate levels of influence by development partners in shaping national and sub-national development agenda priorities (WaterAid 2011)

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 6: Change in complex adaptive systems

5

1 Purpose of this review

This working paper a product of the Triple-S Water Services That Last project 1 is part of an effort to ground IRCrsquos approach to driving and supporting change processes in the water sanitation and hygiene (WASH) sector in the theory and methods of complexity sciences and systems thinking The review of literature from these fields identifies theory rationale and methods underpinning complexity-informed approaches to effecting change in large dynamic complex adaptive systems 2 The insights gained from this body of literature are discussed in relation to the rural water sector in low- and middle-income countriesmdasha sector facing lsquowickedrsquo 3 problems whose solution requires changes in the mindset and behaviours of multiple actors (Rittel and Webber 1973)

The domain of inquiry of this review is the rural water sector in low- and middle-income countries Section 2 describes a series of generally recognisable attributes of the rural water sector as they feature in low- and middle-income countries Salient attributes include the roles and responsibilities of key actors and patterns of interaction between and among key actors institutions and technology in this domain This section also reflects on the multi-faceted and inextricably linked challenges to delivering sustainable services and on the role of international development aid as an external influence in national development agendas and thereby in how public services are financed planned and implemented It reviews the literature on how international development aid can address these challenges by embracing concepts and practices informed by systems thinking and complexity sciences

Section 3 presents concepts and approaches from the complexity sciences and related fields including socio-technical systems as a specific type of complex adaptive system institutional analysis and whole-system change and explores whether they offer new perspectives on alternatives to prevailing national development and water service delivery policy and practice in particular Following the reflection on insights offered by the complexity sciences Section 3 concludes by reconsidering the rural water sector in light of the theory about complex adaptive systems and how change arises in such systems

Section 4 discusses three approaches featured in the literature for their potential to effect systemic change in complex adaptive systems as proposed alternatives to current policy making and planning in international development Given that large-scale systemic change can take several decades Section 5 presents tools for simulating change processes in complex adaptive systems for the purpose of envisaging exploring and experimenting with policy and implementation alternatives Section 6 concludes with a discussion of the insights gained through this review and identifies gaps in the literature about how to foster national systems that can deliver sustainable public services

1 Triple-S Water Services That Last is a six-year multi-country learning initiative to improve water supply to the rural poor that is led by IRC a Netherlands-based mission driven lsquothink and do tankrsquo pursuing the vision of WASH services for everyone forever The initiative is funded by the Bill amp Melinda Gates Foundation and has country programmes active in Ghana and Uganda with smaller initiatives supported in Burkina Faso Mozambique India and Honduras Lessons learned from work in countries feeds up to the international level where Triple-S promotes a re-appraisal of how development assistance to the rural water supply sector is designed and implemented

2 Moriarty Lockwood Carriger and Duti series of four blogposts httpwaterservicesthatlastwordpresscom20140224changing-the-whole-system-to-provide-water-sanitation-and-hygiene-services-that-last MarchndashMay 2014

3 lsquoWickedrsquo describes intractable not-easily-solved problemsmdashsuch as climate change pandemics poverty or natural disastersmdashthat require changes in mindset and behaviours of a great number of actors (Rittel and Webber 1973)

6

WORKING PAPER Change in complex adaptive systems

7

December 2015

In this section the rural water sub-sector and the challenges it faces in providing sustainable services are described Section 21 looks at the administrative and operational arrangements as well as roles and functions of actors involved in rural water sector in low- and middle-income countries Section 22 presents literature about the role and influence of international development as an external change pressure that influences domestic policy financial and technical aspects of rural water services in low- and middle-income countries Section 23 provides an overview of the multiple interlinked and intractable lsquowickedrsquo challenges to delivering services reflecting on the fact that the solutions are not hardware based Section 24 reviews literature arguing that such wicked challenges require alternative approaches grounded in the complexity sciences

21 RURAL WATER SERVICES ndash FROM GOVERNANCE TO SYSTEMS PERSPECTIVES

The overarching purpose of a national WASH sector is to develop and deliver sustainable water sanitation and hygiene services to users for domestic and productive purposes (Plummer and Slaymaker 2007) Although the WASH sector is not clearly demarcated as an entity with precisely identifiable boundaries a number of salient attributes can be identified across different national and regional contexts that enable us to consider the sector a lsquosystemrsquo There are also a number of identifiable activities that people must undertake to ensure that water services are delivered These attributes activities and interactions taken together give rise to a recognisable pattern that when effectively functioning results in the delivery of water services In this section the rural water sector and its salient attributes are described as the domain of focus of this literature review

In terms of attributes there are identifiable legal governance and organisational arrangements in place within nation states that guide the delivery of water services (Rogers and Hall 2003) In general a line ministry or department is the highest mandated authority responsible for ensuring that these services are delivered to all citizens In addition to developing and managing the natural water resource this national authority is tasked with developing and delivering water services to the population for domestic and productive purposes While the national authority for water resources management and

service delivery may also hold the remit for development and delivery of sanitation and hygiene services for the purpose of this review we focus in particular on the legal and administrative arrangements pertaining to rural water service delivery

In the context of decentralised water service delivery models in addition to the national government many other organisations and actors interacting on a continual basis are involved in policy making financial planning and management regulation and service provision activities across multiple administrative levels (Rogers and Hall 2003) Depending upon the national context these actors include sub-national government entities (eg provinces regions districts communes zones municipalities woredas panchayats) water utilities non-governmental organisations (NGOs) users community representatives private operators and capacity-building and financing bodies (de la Harpe 2007) Among this host of actors roles and responsibilities can be differentiated As highlighted by Smits et al (2011) a key distinction is the role of the service authority versus that of the service provider

The service authority generally a government body holds the legal responsibility for service delivery planning coordination regulation and oversight activities as well as technical assistance to water service providers (Smits et al 2011 Lockwood and Smits 2011) In contrast the service provider is the organisation or individual responsible for day-to-day water service which includes the operation maintenance and administration of the water system

How the water service provision role is fulfilled varies widely Most countries have a range of service provision options or service delivery models whereby lsquothe service authority can opt to provide services itself (through a municipal department or municipal company) or hellip delegate this responsibility by contracting an outside agency such as a community-based organisation (CBO) private operator public sector utility or company or non-governmental organisation (NGO) who in turn may hire a private person (plumber or mechanic) to carry out parts of the workrsquo (Smits et al 2011 p5)

In addition to the legal administrative and organisational attributes and interactionsmdashthat is the lsquosocialrsquo components of a water servicemdashthere is also

2 Water services A system description

8

WORKING PAPER Change in complex adaptive systems

the technical component The physical infrastructure required for the delivery of water supply varies greatly among and even within service areas4 depending upon factors such as geographical and hydrological conditions preferences for certain technologies available financial resources and population size and density

Whether the infrastructure is a stand-alone hand pump or a networked gravity-fed piped scheme formal and informal arrangements among the authority provider users civil society and international development organisations are required to ensure sustainable water services (Keohane and Ostrom 1995 Rogers and Hall 2003 Plummer and Slaymaker 2007) These arrangements entail policy- and decision-making processes about responsibilities and actor relationships through which the power responsibilities norms values and formal agreements embedded in laws and policies are negotiated among and implemented by the array of stakeholders whose roles and responsibilities may overlap (Ostrom and Janssen 2004)

This view of interdependent networks of multiple actors or agents interacting across multiple administrative levels embodies a lsquogovernance perspectiversquo (Kooiman 1993 Stoker 1998 DFID 2007) In reflecting on how to provide for a collective interest such as the provision of public services the governance perspective offers a lsquoframework for understanding changing processes of governing characterised by processes of adaptation learning and experimentrsquo (Stoker 1998 p18) This perspective helps to conceptualise how functions related to service delivery are dispersed over a wide array of actors organisations and coordination platforms spanning different national development sectors and administrative levels Notably from a governance perspective while (central) governments continue to play a role in how public services are provided as Bache (2003) notes this role is increasingly one of coordination and steering and is concomitant with an increase in the involvement of non-government actors in policy-making and service delivery The Global Water Partnership (GWP) defines water

governance as lsquothe range of political social economic and administrative systems that are in place to develop and manage water resources and the delivery of water services at different levels of societyrsquo (Rogers and Hall 2003) The literature on governance also highlights the importance of public participation in governance processes for the potential to lsquoimprove the quality of decision making by opening up the decision-making process and making better use of the information and creativity that is available in society improve public under-standing of the management issues at stake make decision making more transparent and might stimulate the different government bodies involved to coordinate their actions more in order to provide serious follow-up to the inputs receivedrsquo and potentially strengthen democratic processes where government does not have all the resources required to lsquomanage an issue effectivelyrsquo (Huitema et al 2009 p5)

The prevailing governance approaches in low- and middle-income countries are context specific and have evolved over long periods of time in response to change pressures such as political social and economic processes (Plummer and Slaymaker 2007) Irrespective of the governance arrangements in a given national context the governance perspective as described here makes it possible to introduce a systems perspective5 to understanding the rural water sector6 This line of inquiry will be addressed in further detail in Section 3

By framing water services as a system that is open to feedback from its environment it becomes possible to consider external change pressures increasingly recognised for their effect on how national development agendasmdashincluding water service deliverymdashare formed and implemented One such pressure is the policy and finance support provided to low- and middle-income countries in the form of international development aid from parties such as UN agencies international finance institutes philanthropic organisations non-governmental organisations and middle- and high-income countries (Rogers and Hall 2003 Mowles et al 2008

4 A service area is the area of jurisdiction and population covered by a service authority Service areas are typically linked to the boundaries of human settlement (towns villages hamlets and scattered rural settlements) but may not correspond precisely with administrative boundaries (IRC Glossary accessed 16 October 2014)

5 A systems perspective takes into account all of the behaviours of a system as a whole in the context of its environment and is lsquoa non-reductionist approach to describing the properties of the system itselfrsquo A description of the whole must include an explanation of the relationships between the parts as well as any additional information needed to describe the behaviour of the entire system (after Bar-Yam 1997 2005)

6 A system is defined by Ryan (2008) as lsquoa representation of an entity as a complex whole open to feedback from its environmentrsquo Ryan (2008) Burke (2006) and other authors on systems and complexity sciences make the important observation that such representations are idealisations based on simplified assumptions Thus although they offer a valid means for identifying and analysing an entity and its dynamics lsquothere are limits to their applicationrsquo (Mowles et al 2008 Ryan 2008)

9

December 2015

Ramalingam 2013) Given the importance of international development aid in shaping national development agenda priorities the next section summarises literature discussing its role and reflects on the dynamic arising in the water sector where it is a main source of finance for service development and a highly influential exogenous change pressure

22 INTERNATIONAL DEVELOPMENT AID EXOGENOUS CHANGE PRESSURE

In this sub-section the role of international development aid in shaping national development agendas of low- and middle-income countries is discussed The international development aid system provides policy financial and technical support to developing nations for education health transportation energy and local and regional economic development and trade A clear understanding of the dynamic created by the involvement of such influential external agents is critical to understanding the current challenges to achieving sustainable water services let alone the realisation of resilient national systems that can develop and deliver the public services required for sustainable and equitable social and economic development (Mowles et al 2008) Many researchers have explored how the current architecture of international development aid delivery is hindering the potential of nations to achieve these social and economic development goals

When developing countries need capital to build infrastructure for public services international development aid actorsmdashranging from development banks funders and bi-lateral government agencies to NGOs and philanthropic organisations (henceforth lsquodevelopment partnersrsquo)mdashmay explicitly partner with a recipient government through policy budget andor technical support strategies and implementation plans (Mowles et al 2008 Ramalingam 2013) It is not uncommon for development partners to bypass nationally led processes and directly implement programmes at the user and community levels (Nimanya et al 2011)

The visions missions and mandates of development partners vary greatly and determine the nature of their interactions with domestic partners including financial investment decisions In addition to investing in infrastructure development partners may support organisational policy and capacity aspects related to the sustainable delivery of a service including both direct support (monitoring maintenance repairs replacements training of staff)

and indirect support (macro-level planning and policy making) These essential components of sustainable water services known as post-construction support or lsquosoftwarersquo are as important as the infrastructure lsquohardwarersquo yet often neglected with actual levels of financial support considered insufficient (Rogers and Hall 2003 Smits et al 2011)

Reasons for this neglect include the desire to focus resources on increasing coverage rates for unserved populations (WHO 2012) perceptions about the risk of corruption faulty assumptions about the lsquobestrsquo governance arrangements for post-construction activities (Schouten and Moriarty 2003) and the desire to see tangible easily measurable results from an investment (Garandeau et al 2009)

Understanding the sources of financing provides insight into how WASH policy priorities are determined The Organisation for Economic Co-operation and Development (OECD 2009) distinguishes the sources of financing for WASH services as the three Ts tariffs which are contributions paid by service users transfers in the form of assistance from development partners and taxes which are levied by national or regional governments In many countries the funding from international aid (transfers to use the OECD nomenclature) is at least as much as the funding from the two domestic sources tariffs and taxes (Figure 1)

Financing from transfers is not problematic on its own and in fact does tremendous good in many countries However since lsquohe who pays the piper calls the tunersquo national policy strategy and governance reform interventions are frequently and significantly influenced by development partnersrsquo priorities (Water Aid 2011) especially where transfers are collectively greater than domestic sources of financing generated through taxes and tariffs and where transfers are made outside the national policy agenda

As Figure 1 shows lsquodonor aid to the WASH sector as a percentage of GDP is higher than government budget allocations for WASH in Cambodia Ghana Liberia Madagascar Rwanda Timor-Leste and Uganda indicating both a donor-dominated sector and also that significant amounts of aid to the WASH sector in these countries is not recorded in central government budgets and accounts or is off-budgetrsquo (WaterAid 2011 p35) This disparity in international and domestic funding translates into disproportionate levels of influence by development partners in shaping national and sub-national development agenda priorities (WaterAid 2011)

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 7: Change in complex adaptive systems

6

WORKING PAPER Change in complex adaptive systems

7

December 2015

In this section the rural water sub-sector and the challenges it faces in providing sustainable services are described Section 21 looks at the administrative and operational arrangements as well as roles and functions of actors involved in rural water sector in low- and middle-income countries Section 22 presents literature about the role and influence of international development as an external change pressure that influences domestic policy financial and technical aspects of rural water services in low- and middle-income countries Section 23 provides an overview of the multiple interlinked and intractable lsquowickedrsquo challenges to delivering services reflecting on the fact that the solutions are not hardware based Section 24 reviews literature arguing that such wicked challenges require alternative approaches grounded in the complexity sciences

21 RURAL WATER SERVICES ndash FROM GOVERNANCE TO SYSTEMS PERSPECTIVES

The overarching purpose of a national WASH sector is to develop and deliver sustainable water sanitation and hygiene services to users for domestic and productive purposes (Plummer and Slaymaker 2007) Although the WASH sector is not clearly demarcated as an entity with precisely identifiable boundaries a number of salient attributes can be identified across different national and regional contexts that enable us to consider the sector a lsquosystemrsquo There are also a number of identifiable activities that people must undertake to ensure that water services are delivered These attributes activities and interactions taken together give rise to a recognisable pattern that when effectively functioning results in the delivery of water services In this section the rural water sector and its salient attributes are described as the domain of focus of this literature review

In terms of attributes there are identifiable legal governance and organisational arrangements in place within nation states that guide the delivery of water services (Rogers and Hall 2003) In general a line ministry or department is the highest mandated authority responsible for ensuring that these services are delivered to all citizens In addition to developing and managing the natural water resource this national authority is tasked with developing and delivering water services to the population for domestic and productive purposes While the national authority for water resources management and

service delivery may also hold the remit for development and delivery of sanitation and hygiene services for the purpose of this review we focus in particular on the legal and administrative arrangements pertaining to rural water service delivery

In the context of decentralised water service delivery models in addition to the national government many other organisations and actors interacting on a continual basis are involved in policy making financial planning and management regulation and service provision activities across multiple administrative levels (Rogers and Hall 2003) Depending upon the national context these actors include sub-national government entities (eg provinces regions districts communes zones municipalities woredas panchayats) water utilities non-governmental organisations (NGOs) users community representatives private operators and capacity-building and financing bodies (de la Harpe 2007) Among this host of actors roles and responsibilities can be differentiated As highlighted by Smits et al (2011) a key distinction is the role of the service authority versus that of the service provider

The service authority generally a government body holds the legal responsibility for service delivery planning coordination regulation and oversight activities as well as technical assistance to water service providers (Smits et al 2011 Lockwood and Smits 2011) In contrast the service provider is the organisation or individual responsible for day-to-day water service which includes the operation maintenance and administration of the water system

How the water service provision role is fulfilled varies widely Most countries have a range of service provision options or service delivery models whereby lsquothe service authority can opt to provide services itself (through a municipal department or municipal company) or hellip delegate this responsibility by contracting an outside agency such as a community-based organisation (CBO) private operator public sector utility or company or non-governmental organisation (NGO) who in turn may hire a private person (plumber or mechanic) to carry out parts of the workrsquo (Smits et al 2011 p5)

In addition to the legal administrative and organisational attributes and interactionsmdashthat is the lsquosocialrsquo components of a water servicemdashthere is also

2 Water services A system description

8

WORKING PAPER Change in complex adaptive systems

the technical component The physical infrastructure required for the delivery of water supply varies greatly among and even within service areas4 depending upon factors such as geographical and hydrological conditions preferences for certain technologies available financial resources and population size and density

Whether the infrastructure is a stand-alone hand pump or a networked gravity-fed piped scheme formal and informal arrangements among the authority provider users civil society and international development organisations are required to ensure sustainable water services (Keohane and Ostrom 1995 Rogers and Hall 2003 Plummer and Slaymaker 2007) These arrangements entail policy- and decision-making processes about responsibilities and actor relationships through which the power responsibilities norms values and formal agreements embedded in laws and policies are negotiated among and implemented by the array of stakeholders whose roles and responsibilities may overlap (Ostrom and Janssen 2004)

This view of interdependent networks of multiple actors or agents interacting across multiple administrative levels embodies a lsquogovernance perspectiversquo (Kooiman 1993 Stoker 1998 DFID 2007) In reflecting on how to provide for a collective interest such as the provision of public services the governance perspective offers a lsquoframework for understanding changing processes of governing characterised by processes of adaptation learning and experimentrsquo (Stoker 1998 p18) This perspective helps to conceptualise how functions related to service delivery are dispersed over a wide array of actors organisations and coordination platforms spanning different national development sectors and administrative levels Notably from a governance perspective while (central) governments continue to play a role in how public services are provided as Bache (2003) notes this role is increasingly one of coordination and steering and is concomitant with an increase in the involvement of non-government actors in policy-making and service delivery The Global Water Partnership (GWP) defines water

governance as lsquothe range of political social economic and administrative systems that are in place to develop and manage water resources and the delivery of water services at different levels of societyrsquo (Rogers and Hall 2003) The literature on governance also highlights the importance of public participation in governance processes for the potential to lsquoimprove the quality of decision making by opening up the decision-making process and making better use of the information and creativity that is available in society improve public under-standing of the management issues at stake make decision making more transparent and might stimulate the different government bodies involved to coordinate their actions more in order to provide serious follow-up to the inputs receivedrsquo and potentially strengthen democratic processes where government does not have all the resources required to lsquomanage an issue effectivelyrsquo (Huitema et al 2009 p5)

The prevailing governance approaches in low- and middle-income countries are context specific and have evolved over long periods of time in response to change pressures such as political social and economic processes (Plummer and Slaymaker 2007) Irrespective of the governance arrangements in a given national context the governance perspective as described here makes it possible to introduce a systems perspective5 to understanding the rural water sector6 This line of inquiry will be addressed in further detail in Section 3

By framing water services as a system that is open to feedback from its environment it becomes possible to consider external change pressures increasingly recognised for their effect on how national development agendasmdashincluding water service deliverymdashare formed and implemented One such pressure is the policy and finance support provided to low- and middle-income countries in the form of international development aid from parties such as UN agencies international finance institutes philanthropic organisations non-governmental organisations and middle- and high-income countries (Rogers and Hall 2003 Mowles et al 2008

4 A service area is the area of jurisdiction and population covered by a service authority Service areas are typically linked to the boundaries of human settlement (towns villages hamlets and scattered rural settlements) but may not correspond precisely with administrative boundaries (IRC Glossary accessed 16 October 2014)

5 A systems perspective takes into account all of the behaviours of a system as a whole in the context of its environment and is lsquoa non-reductionist approach to describing the properties of the system itselfrsquo A description of the whole must include an explanation of the relationships between the parts as well as any additional information needed to describe the behaviour of the entire system (after Bar-Yam 1997 2005)

6 A system is defined by Ryan (2008) as lsquoa representation of an entity as a complex whole open to feedback from its environmentrsquo Ryan (2008) Burke (2006) and other authors on systems and complexity sciences make the important observation that such representations are idealisations based on simplified assumptions Thus although they offer a valid means for identifying and analysing an entity and its dynamics lsquothere are limits to their applicationrsquo (Mowles et al 2008 Ryan 2008)

9

December 2015

Ramalingam 2013) Given the importance of international development aid in shaping national development agenda priorities the next section summarises literature discussing its role and reflects on the dynamic arising in the water sector where it is a main source of finance for service development and a highly influential exogenous change pressure

22 INTERNATIONAL DEVELOPMENT AID EXOGENOUS CHANGE PRESSURE

In this sub-section the role of international development aid in shaping national development agendas of low- and middle-income countries is discussed The international development aid system provides policy financial and technical support to developing nations for education health transportation energy and local and regional economic development and trade A clear understanding of the dynamic created by the involvement of such influential external agents is critical to understanding the current challenges to achieving sustainable water services let alone the realisation of resilient national systems that can develop and deliver the public services required for sustainable and equitable social and economic development (Mowles et al 2008) Many researchers have explored how the current architecture of international development aid delivery is hindering the potential of nations to achieve these social and economic development goals

When developing countries need capital to build infrastructure for public services international development aid actorsmdashranging from development banks funders and bi-lateral government agencies to NGOs and philanthropic organisations (henceforth lsquodevelopment partnersrsquo)mdashmay explicitly partner with a recipient government through policy budget andor technical support strategies and implementation plans (Mowles et al 2008 Ramalingam 2013) It is not uncommon for development partners to bypass nationally led processes and directly implement programmes at the user and community levels (Nimanya et al 2011)

The visions missions and mandates of development partners vary greatly and determine the nature of their interactions with domestic partners including financial investment decisions In addition to investing in infrastructure development partners may support organisational policy and capacity aspects related to the sustainable delivery of a service including both direct support (monitoring maintenance repairs replacements training of staff)

and indirect support (macro-level planning and policy making) These essential components of sustainable water services known as post-construction support or lsquosoftwarersquo are as important as the infrastructure lsquohardwarersquo yet often neglected with actual levels of financial support considered insufficient (Rogers and Hall 2003 Smits et al 2011)

Reasons for this neglect include the desire to focus resources on increasing coverage rates for unserved populations (WHO 2012) perceptions about the risk of corruption faulty assumptions about the lsquobestrsquo governance arrangements for post-construction activities (Schouten and Moriarty 2003) and the desire to see tangible easily measurable results from an investment (Garandeau et al 2009)

Understanding the sources of financing provides insight into how WASH policy priorities are determined The Organisation for Economic Co-operation and Development (OECD 2009) distinguishes the sources of financing for WASH services as the three Ts tariffs which are contributions paid by service users transfers in the form of assistance from development partners and taxes which are levied by national or regional governments In many countries the funding from international aid (transfers to use the OECD nomenclature) is at least as much as the funding from the two domestic sources tariffs and taxes (Figure 1)

Financing from transfers is not problematic on its own and in fact does tremendous good in many countries However since lsquohe who pays the piper calls the tunersquo national policy strategy and governance reform interventions are frequently and significantly influenced by development partnersrsquo priorities (Water Aid 2011) especially where transfers are collectively greater than domestic sources of financing generated through taxes and tariffs and where transfers are made outside the national policy agenda

As Figure 1 shows lsquodonor aid to the WASH sector as a percentage of GDP is higher than government budget allocations for WASH in Cambodia Ghana Liberia Madagascar Rwanda Timor-Leste and Uganda indicating both a donor-dominated sector and also that significant amounts of aid to the WASH sector in these countries is not recorded in central government budgets and accounts or is off-budgetrsquo (WaterAid 2011 p35) This disparity in international and domestic funding translates into disproportionate levels of influence by development partners in shaping national and sub-national development agenda priorities (WaterAid 2011)

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 8: Change in complex adaptive systems

7

December 2015

In this section the rural water sub-sector and the challenges it faces in providing sustainable services are described Section 21 looks at the administrative and operational arrangements as well as roles and functions of actors involved in rural water sector in low- and middle-income countries Section 22 presents literature about the role and influence of international development as an external change pressure that influences domestic policy financial and technical aspects of rural water services in low- and middle-income countries Section 23 provides an overview of the multiple interlinked and intractable lsquowickedrsquo challenges to delivering services reflecting on the fact that the solutions are not hardware based Section 24 reviews literature arguing that such wicked challenges require alternative approaches grounded in the complexity sciences

21 RURAL WATER SERVICES ndash FROM GOVERNANCE TO SYSTEMS PERSPECTIVES

The overarching purpose of a national WASH sector is to develop and deliver sustainable water sanitation and hygiene services to users for domestic and productive purposes (Plummer and Slaymaker 2007) Although the WASH sector is not clearly demarcated as an entity with precisely identifiable boundaries a number of salient attributes can be identified across different national and regional contexts that enable us to consider the sector a lsquosystemrsquo There are also a number of identifiable activities that people must undertake to ensure that water services are delivered These attributes activities and interactions taken together give rise to a recognisable pattern that when effectively functioning results in the delivery of water services In this section the rural water sector and its salient attributes are described as the domain of focus of this literature review

In terms of attributes there are identifiable legal governance and organisational arrangements in place within nation states that guide the delivery of water services (Rogers and Hall 2003) In general a line ministry or department is the highest mandated authority responsible for ensuring that these services are delivered to all citizens In addition to developing and managing the natural water resource this national authority is tasked with developing and delivering water services to the population for domestic and productive purposes While the national authority for water resources management and

service delivery may also hold the remit for development and delivery of sanitation and hygiene services for the purpose of this review we focus in particular on the legal and administrative arrangements pertaining to rural water service delivery

In the context of decentralised water service delivery models in addition to the national government many other organisations and actors interacting on a continual basis are involved in policy making financial planning and management regulation and service provision activities across multiple administrative levels (Rogers and Hall 2003) Depending upon the national context these actors include sub-national government entities (eg provinces regions districts communes zones municipalities woredas panchayats) water utilities non-governmental organisations (NGOs) users community representatives private operators and capacity-building and financing bodies (de la Harpe 2007) Among this host of actors roles and responsibilities can be differentiated As highlighted by Smits et al (2011) a key distinction is the role of the service authority versus that of the service provider

The service authority generally a government body holds the legal responsibility for service delivery planning coordination regulation and oversight activities as well as technical assistance to water service providers (Smits et al 2011 Lockwood and Smits 2011) In contrast the service provider is the organisation or individual responsible for day-to-day water service which includes the operation maintenance and administration of the water system

How the water service provision role is fulfilled varies widely Most countries have a range of service provision options or service delivery models whereby lsquothe service authority can opt to provide services itself (through a municipal department or municipal company) or hellip delegate this responsibility by contracting an outside agency such as a community-based organisation (CBO) private operator public sector utility or company or non-governmental organisation (NGO) who in turn may hire a private person (plumber or mechanic) to carry out parts of the workrsquo (Smits et al 2011 p5)

In addition to the legal administrative and organisational attributes and interactionsmdashthat is the lsquosocialrsquo components of a water servicemdashthere is also

2 Water services A system description

8

WORKING PAPER Change in complex adaptive systems

the technical component The physical infrastructure required for the delivery of water supply varies greatly among and even within service areas4 depending upon factors such as geographical and hydrological conditions preferences for certain technologies available financial resources and population size and density

Whether the infrastructure is a stand-alone hand pump or a networked gravity-fed piped scheme formal and informal arrangements among the authority provider users civil society and international development organisations are required to ensure sustainable water services (Keohane and Ostrom 1995 Rogers and Hall 2003 Plummer and Slaymaker 2007) These arrangements entail policy- and decision-making processes about responsibilities and actor relationships through which the power responsibilities norms values and formal agreements embedded in laws and policies are negotiated among and implemented by the array of stakeholders whose roles and responsibilities may overlap (Ostrom and Janssen 2004)

This view of interdependent networks of multiple actors or agents interacting across multiple administrative levels embodies a lsquogovernance perspectiversquo (Kooiman 1993 Stoker 1998 DFID 2007) In reflecting on how to provide for a collective interest such as the provision of public services the governance perspective offers a lsquoframework for understanding changing processes of governing characterised by processes of adaptation learning and experimentrsquo (Stoker 1998 p18) This perspective helps to conceptualise how functions related to service delivery are dispersed over a wide array of actors organisations and coordination platforms spanning different national development sectors and administrative levels Notably from a governance perspective while (central) governments continue to play a role in how public services are provided as Bache (2003) notes this role is increasingly one of coordination and steering and is concomitant with an increase in the involvement of non-government actors in policy-making and service delivery The Global Water Partnership (GWP) defines water

governance as lsquothe range of political social economic and administrative systems that are in place to develop and manage water resources and the delivery of water services at different levels of societyrsquo (Rogers and Hall 2003) The literature on governance also highlights the importance of public participation in governance processes for the potential to lsquoimprove the quality of decision making by opening up the decision-making process and making better use of the information and creativity that is available in society improve public under-standing of the management issues at stake make decision making more transparent and might stimulate the different government bodies involved to coordinate their actions more in order to provide serious follow-up to the inputs receivedrsquo and potentially strengthen democratic processes where government does not have all the resources required to lsquomanage an issue effectivelyrsquo (Huitema et al 2009 p5)

The prevailing governance approaches in low- and middle-income countries are context specific and have evolved over long periods of time in response to change pressures such as political social and economic processes (Plummer and Slaymaker 2007) Irrespective of the governance arrangements in a given national context the governance perspective as described here makes it possible to introduce a systems perspective5 to understanding the rural water sector6 This line of inquiry will be addressed in further detail in Section 3

By framing water services as a system that is open to feedback from its environment it becomes possible to consider external change pressures increasingly recognised for their effect on how national development agendasmdashincluding water service deliverymdashare formed and implemented One such pressure is the policy and finance support provided to low- and middle-income countries in the form of international development aid from parties such as UN agencies international finance institutes philanthropic organisations non-governmental organisations and middle- and high-income countries (Rogers and Hall 2003 Mowles et al 2008

4 A service area is the area of jurisdiction and population covered by a service authority Service areas are typically linked to the boundaries of human settlement (towns villages hamlets and scattered rural settlements) but may not correspond precisely with administrative boundaries (IRC Glossary accessed 16 October 2014)

5 A systems perspective takes into account all of the behaviours of a system as a whole in the context of its environment and is lsquoa non-reductionist approach to describing the properties of the system itselfrsquo A description of the whole must include an explanation of the relationships between the parts as well as any additional information needed to describe the behaviour of the entire system (after Bar-Yam 1997 2005)

6 A system is defined by Ryan (2008) as lsquoa representation of an entity as a complex whole open to feedback from its environmentrsquo Ryan (2008) Burke (2006) and other authors on systems and complexity sciences make the important observation that such representations are idealisations based on simplified assumptions Thus although they offer a valid means for identifying and analysing an entity and its dynamics lsquothere are limits to their applicationrsquo (Mowles et al 2008 Ryan 2008)

9

December 2015

Ramalingam 2013) Given the importance of international development aid in shaping national development agenda priorities the next section summarises literature discussing its role and reflects on the dynamic arising in the water sector where it is a main source of finance for service development and a highly influential exogenous change pressure

22 INTERNATIONAL DEVELOPMENT AID EXOGENOUS CHANGE PRESSURE

In this sub-section the role of international development aid in shaping national development agendas of low- and middle-income countries is discussed The international development aid system provides policy financial and technical support to developing nations for education health transportation energy and local and regional economic development and trade A clear understanding of the dynamic created by the involvement of such influential external agents is critical to understanding the current challenges to achieving sustainable water services let alone the realisation of resilient national systems that can develop and deliver the public services required for sustainable and equitable social and economic development (Mowles et al 2008) Many researchers have explored how the current architecture of international development aid delivery is hindering the potential of nations to achieve these social and economic development goals

When developing countries need capital to build infrastructure for public services international development aid actorsmdashranging from development banks funders and bi-lateral government agencies to NGOs and philanthropic organisations (henceforth lsquodevelopment partnersrsquo)mdashmay explicitly partner with a recipient government through policy budget andor technical support strategies and implementation plans (Mowles et al 2008 Ramalingam 2013) It is not uncommon for development partners to bypass nationally led processes and directly implement programmes at the user and community levels (Nimanya et al 2011)

The visions missions and mandates of development partners vary greatly and determine the nature of their interactions with domestic partners including financial investment decisions In addition to investing in infrastructure development partners may support organisational policy and capacity aspects related to the sustainable delivery of a service including both direct support (monitoring maintenance repairs replacements training of staff)

and indirect support (macro-level planning and policy making) These essential components of sustainable water services known as post-construction support or lsquosoftwarersquo are as important as the infrastructure lsquohardwarersquo yet often neglected with actual levels of financial support considered insufficient (Rogers and Hall 2003 Smits et al 2011)

Reasons for this neglect include the desire to focus resources on increasing coverage rates for unserved populations (WHO 2012) perceptions about the risk of corruption faulty assumptions about the lsquobestrsquo governance arrangements for post-construction activities (Schouten and Moriarty 2003) and the desire to see tangible easily measurable results from an investment (Garandeau et al 2009)

Understanding the sources of financing provides insight into how WASH policy priorities are determined The Organisation for Economic Co-operation and Development (OECD 2009) distinguishes the sources of financing for WASH services as the three Ts tariffs which are contributions paid by service users transfers in the form of assistance from development partners and taxes which are levied by national or regional governments In many countries the funding from international aid (transfers to use the OECD nomenclature) is at least as much as the funding from the two domestic sources tariffs and taxes (Figure 1)

Financing from transfers is not problematic on its own and in fact does tremendous good in many countries However since lsquohe who pays the piper calls the tunersquo national policy strategy and governance reform interventions are frequently and significantly influenced by development partnersrsquo priorities (Water Aid 2011) especially where transfers are collectively greater than domestic sources of financing generated through taxes and tariffs and where transfers are made outside the national policy agenda

As Figure 1 shows lsquodonor aid to the WASH sector as a percentage of GDP is higher than government budget allocations for WASH in Cambodia Ghana Liberia Madagascar Rwanda Timor-Leste and Uganda indicating both a donor-dominated sector and also that significant amounts of aid to the WASH sector in these countries is not recorded in central government budgets and accounts or is off-budgetrsquo (WaterAid 2011 p35) This disparity in international and domestic funding translates into disproportionate levels of influence by development partners in shaping national and sub-national development agenda priorities (WaterAid 2011)

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 9: Change in complex adaptive systems

8

WORKING PAPER Change in complex adaptive systems

the technical component The physical infrastructure required for the delivery of water supply varies greatly among and even within service areas4 depending upon factors such as geographical and hydrological conditions preferences for certain technologies available financial resources and population size and density

Whether the infrastructure is a stand-alone hand pump or a networked gravity-fed piped scheme formal and informal arrangements among the authority provider users civil society and international development organisations are required to ensure sustainable water services (Keohane and Ostrom 1995 Rogers and Hall 2003 Plummer and Slaymaker 2007) These arrangements entail policy- and decision-making processes about responsibilities and actor relationships through which the power responsibilities norms values and formal agreements embedded in laws and policies are negotiated among and implemented by the array of stakeholders whose roles and responsibilities may overlap (Ostrom and Janssen 2004)

This view of interdependent networks of multiple actors or agents interacting across multiple administrative levels embodies a lsquogovernance perspectiversquo (Kooiman 1993 Stoker 1998 DFID 2007) In reflecting on how to provide for a collective interest such as the provision of public services the governance perspective offers a lsquoframework for understanding changing processes of governing characterised by processes of adaptation learning and experimentrsquo (Stoker 1998 p18) This perspective helps to conceptualise how functions related to service delivery are dispersed over a wide array of actors organisations and coordination platforms spanning different national development sectors and administrative levels Notably from a governance perspective while (central) governments continue to play a role in how public services are provided as Bache (2003) notes this role is increasingly one of coordination and steering and is concomitant with an increase in the involvement of non-government actors in policy-making and service delivery The Global Water Partnership (GWP) defines water

governance as lsquothe range of political social economic and administrative systems that are in place to develop and manage water resources and the delivery of water services at different levels of societyrsquo (Rogers and Hall 2003) The literature on governance also highlights the importance of public participation in governance processes for the potential to lsquoimprove the quality of decision making by opening up the decision-making process and making better use of the information and creativity that is available in society improve public under-standing of the management issues at stake make decision making more transparent and might stimulate the different government bodies involved to coordinate their actions more in order to provide serious follow-up to the inputs receivedrsquo and potentially strengthen democratic processes where government does not have all the resources required to lsquomanage an issue effectivelyrsquo (Huitema et al 2009 p5)

The prevailing governance approaches in low- and middle-income countries are context specific and have evolved over long periods of time in response to change pressures such as political social and economic processes (Plummer and Slaymaker 2007) Irrespective of the governance arrangements in a given national context the governance perspective as described here makes it possible to introduce a systems perspective5 to understanding the rural water sector6 This line of inquiry will be addressed in further detail in Section 3

By framing water services as a system that is open to feedback from its environment it becomes possible to consider external change pressures increasingly recognised for their effect on how national development agendasmdashincluding water service deliverymdashare formed and implemented One such pressure is the policy and finance support provided to low- and middle-income countries in the form of international development aid from parties such as UN agencies international finance institutes philanthropic organisations non-governmental organisations and middle- and high-income countries (Rogers and Hall 2003 Mowles et al 2008

4 A service area is the area of jurisdiction and population covered by a service authority Service areas are typically linked to the boundaries of human settlement (towns villages hamlets and scattered rural settlements) but may not correspond precisely with administrative boundaries (IRC Glossary accessed 16 October 2014)

5 A systems perspective takes into account all of the behaviours of a system as a whole in the context of its environment and is lsquoa non-reductionist approach to describing the properties of the system itselfrsquo A description of the whole must include an explanation of the relationships between the parts as well as any additional information needed to describe the behaviour of the entire system (after Bar-Yam 1997 2005)

6 A system is defined by Ryan (2008) as lsquoa representation of an entity as a complex whole open to feedback from its environmentrsquo Ryan (2008) Burke (2006) and other authors on systems and complexity sciences make the important observation that such representations are idealisations based on simplified assumptions Thus although they offer a valid means for identifying and analysing an entity and its dynamics lsquothere are limits to their applicationrsquo (Mowles et al 2008 Ryan 2008)

9

December 2015

Ramalingam 2013) Given the importance of international development aid in shaping national development agenda priorities the next section summarises literature discussing its role and reflects on the dynamic arising in the water sector where it is a main source of finance for service development and a highly influential exogenous change pressure

22 INTERNATIONAL DEVELOPMENT AID EXOGENOUS CHANGE PRESSURE

In this sub-section the role of international development aid in shaping national development agendas of low- and middle-income countries is discussed The international development aid system provides policy financial and technical support to developing nations for education health transportation energy and local and regional economic development and trade A clear understanding of the dynamic created by the involvement of such influential external agents is critical to understanding the current challenges to achieving sustainable water services let alone the realisation of resilient national systems that can develop and deliver the public services required for sustainable and equitable social and economic development (Mowles et al 2008) Many researchers have explored how the current architecture of international development aid delivery is hindering the potential of nations to achieve these social and economic development goals

When developing countries need capital to build infrastructure for public services international development aid actorsmdashranging from development banks funders and bi-lateral government agencies to NGOs and philanthropic organisations (henceforth lsquodevelopment partnersrsquo)mdashmay explicitly partner with a recipient government through policy budget andor technical support strategies and implementation plans (Mowles et al 2008 Ramalingam 2013) It is not uncommon for development partners to bypass nationally led processes and directly implement programmes at the user and community levels (Nimanya et al 2011)

The visions missions and mandates of development partners vary greatly and determine the nature of their interactions with domestic partners including financial investment decisions In addition to investing in infrastructure development partners may support organisational policy and capacity aspects related to the sustainable delivery of a service including both direct support (monitoring maintenance repairs replacements training of staff)

and indirect support (macro-level planning and policy making) These essential components of sustainable water services known as post-construction support or lsquosoftwarersquo are as important as the infrastructure lsquohardwarersquo yet often neglected with actual levels of financial support considered insufficient (Rogers and Hall 2003 Smits et al 2011)

Reasons for this neglect include the desire to focus resources on increasing coverage rates for unserved populations (WHO 2012) perceptions about the risk of corruption faulty assumptions about the lsquobestrsquo governance arrangements for post-construction activities (Schouten and Moriarty 2003) and the desire to see tangible easily measurable results from an investment (Garandeau et al 2009)

Understanding the sources of financing provides insight into how WASH policy priorities are determined The Organisation for Economic Co-operation and Development (OECD 2009) distinguishes the sources of financing for WASH services as the three Ts tariffs which are contributions paid by service users transfers in the form of assistance from development partners and taxes which are levied by national or regional governments In many countries the funding from international aid (transfers to use the OECD nomenclature) is at least as much as the funding from the two domestic sources tariffs and taxes (Figure 1)

Financing from transfers is not problematic on its own and in fact does tremendous good in many countries However since lsquohe who pays the piper calls the tunersquo national policy strategy and governance reform interventions are frequently and significantly influenced by development partnersrsquo priorities (Water Aid 2011) especially where transfers are collectively greater than domestic sources of financing generated through taxes and tariffs and where transfers are made outside the national policy agenda

As Figure 1 shows lsquodonor aid to the WASH sector as a percentage of GDP is higher than government budget allocations for WASH in Cambodia Ghana Liberia Madagascar Rwanda Timor-Leste and Uganda indicating both a donor-dominated sector and also that significant amounts of aid to the WASH sector in these countries is not recorded in central government budgets and accounts or is off-budgetrsquo (WaterAid 2011 p35) This disparity in international and domestic funding translates into disproportionate levels of influence by development partners in shaping national and sub-national development agenda priorities (WaterAid 2011)

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 10: Change in complex adaptive systems

9

December 2015

Ramalingam 2013) Given the importance of international development aid in shaping national development agenda priorities the next section summarises literature discussing its role and reflects on the dynamic arising in the water sector where it is a main source of finance for service development and a highly influential exogenous change pressure

22 INTERNATIONAL DEVELOPMENT AID EXOGENOUS CHANGE PRESSURE

In this sub-section the role of international development aid in shaping national development agendas of low- and middle-income countries is discussed The international development aid system provides policy financial and technical support to developing nations for education health transportation energy and local and regional economic development and trade A clear understanding of the dynamic created by the involvement of such influential external agents is critical to understanding the current challenges to achieving sustainable water services let alone the realisation of resilient national systems that can develop and deliver the public services required for sustainable and equitable social and economic development (Mowles et al 2008) Many researchers have explored how the current architecture of international development aid delivery is hindering the potential of nations to achieve these social and economic development goals

When developing countries need capital to build infrastructure for public services international development aid actorsmdashranging from development banks funders and bi-lateral government agencies to NGOs and philanthropic organisations (henceforth lsquodevelopment partnersrsquo)mdashmay explicitly partner with a recipient government through policy budget andor technical support strategies and implementation plans (Mowles et al 2008 Ramalingam 2013) It is not uncommon for development partners to bypass nationally led processes and directly implement programmes at the user and community levels (Nimanya et al 2011)

The visions missions and mandates of development partners vary greatly and determine the nature of their interactions with domestic partners including financial investment decisions In addition to investing in infrastructure development partners may support organisational policy and capacity aspects related to the sustainable delivery of a service including both direct support (monitoring maintenance repairs replacements training of staff)

and indirect support (macro-level planning and policy making) These essential components of sustainable water services known as post-construction support or lsquosoftwarersquo are as important as the infrastructure lsquohardwarersquo yet often neglected with actual levels of financial support considered insufficient (Rogers and Hall 2003 Smits et al 2011)

Reasons for this neglect include the desire to focus resources on increasing coverage rates for unserved populations (WHO 2012) perceptions about the risk of corruption faulty assumptions about the lsquobestrsquo governance arrangements for post-construction activities (Schouten and Moriarty 2003) and the desire to see tangible easily measurable results from an investment (Garandeau et al 2009)

Understanding the sources of financing provides insight into how WASH policy priorities are determined The Organisation for Economic Co-operation and Development (OECD 2009) distinguishes the sources of financing for WASH services as the three Ts tariffs which are contributions paid by service users transfers in the form of assistance from development partners and taxes which are levied by national or regional governments In many countries the funding from international aid (transfers to use the OECD nomenclature) is at least as much as the funding from the two domestic sources tariffs and taxes (Figure 1)

Financing from transfers is not problematic on its own and in fact does tremendous good in many countries However since lsquohe who pays the piper calls the tunersquo national policy strategy and governance reform interventions are frequently and significantly influenced by development partnersrsquo priorities (Water Aid 2011) especially where transfers are collectively greater than domestic sources of financing generated through taxes and tariffs and where transfers are made outside the national policy agenda

As Figure 1 shows lsquodonor aid to the WASH sector as a percentage of GDP is higher than government budget allocations for WASH in Cambodia Ghana Liberia Madagascar Rwanda Timor-Leste and Uganda indicating both a donor-dominated sector and also that significant amounts of aid to the WASH sector in these countries is not recorded in central government budgets and accounts or is off-budgetrsquo (WaterAid 2011 p35) This disparity in international and domestic funding translates into disproportionate levels of influence by development partners in shaping national and sub-national development agenda priorities (WaterAid 2011)

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 11: Change in complex adaptive systems

10

WORKING PAPER Change in complex adaptive systems

Development partners are a heterogeneous group in terms of organisational visions missions and approaches to providing development aid resources Some operate with and through national policy budget and coordination processes others work lsquooff-budgetrsquo and may provide a significant proportion of investmentmdash30 by one estimation in the rural sub-sector in Uganda (Nimanya et al 2011) Development partners are not democratically elected entities yet their aid to national WASH sectors exceeds domestic sources of financing from mandated public authorities (WaterAid 2011) Their influence must be accounted for when seeking to understand how systemic change can occur No rigorous comparisons have been made of how WASH sector policies and outcomes differ between countries based on the proportion of domestic funding to transfers

Development partners have also made well-intended interventions to strengthen governance foster resilient national systems and build sector capacity often by introducing governance structures based on examples of more or less effective national systems in high-income countries This has been called lsquosystemic isomorphic mimicryrsquo a concept borrowed from the natural sciences where it refers to a species that evolves to resemble the form of another species without its functions (eg a fly that evolves to look like a bee to avoid predation but lacks the beersquos protection mechanism of a toxic sting) (Pritchett et al 2010 Andrews et al 2013) In governance and policy reform imitation to address lsquocapability trapsrsquo is problematic (Pritchett et al 2010)

A copy-and-paste approach to implementing large-scale policy and organisational reform in one socio-

7 Institutions are rules that are accepted by all those involved are used in practice and have some sort of durability (Ghorbani 2013)

Source Water Aid 2011ODA = overseas development aid

FIGURE 1 DEVELOPMENT PARTNER FUNDING AND NATIONAL ALLOCATIONS TO WASH AS PERCENTAGE OF GDP

Angola

Bangladesh

Burkina Faso

Cambodia

Central African

Republic

Cote dlvoire

Ethiopia

India

Kenya

Lesotho

Liberia

Madagascar

Malawi

Mali

Mozambique

Nepal

Niger

Nigeria

Pakistan

Papua New Guinea

Rwanda

Senegal

Sierra Leone

Tanzania

Timor Leste

Uganda

Zambia

00 05 10 15 20 25 30

GovernmentWashallocationas GDP

ODA allocatedto WASHas GDP

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 12: Change in complex adaptive systems

11

December 2015

technical system based on best practices from another setting rarely produces the desired results because the two settingsrsquo policy and organisational environments evolved through different social political economic and technical selection pressures (Pritchett et al 2010 Andrews et al 2013) Moreover this approach to policy and governance reform undercuts lsquoindigenous learning the legitimacy of change and the support of key political constituenciesrsquo (Pritchett et al 2013a p1)

A third unintended effect of international development aid arises from the conventional three- to five-year duration of development interventions even for large-scale institutional change initiatives Williamson (2000) reflecting on the current state and future offerings of new institutional economics finds that the rate of change differs by the level or type of institution7 (Figure 2)

For informal institutions (norms and culture) change occurs every 100 to 1000 years Change to formal rules (laws and regulations) requires 10 to 100 years Agreements and contracts change in one to 10 years At the lowest level operational rules change continually (Williamson 2000 Ghorbani et al 2010 Van Tongeren 2014)

In other words institutional and governance systems require time to develop agents within those systems must internalise change and identify their changing roles in the evolving system The resulting change is an outcome of domestic and possibly also international social political economic and increasingly environmental and resource use pressures (Huitema et al 2009) Acquiring new functions within a system requires not just financial resources but also the time and space to learn from trial and error So too national systems for the delivery of services have evolved in context-specific ways over long periods in response to political social and economic processes (Plummer and Slaymaker 2007) Williamson (2000) considers not only institutional levels and their change frequency but also lsquodesign opportunitiesrsquo for policy makers to achieve change in formal rules (first-order economizing) play-of-game rules (second-order economizing) and contractual relations or lsquoprivate orderingrsquo (third-order economizing) (Williamson 2000 pp598ndash99)

International development aid is recognised as beneficial in supporting recipient nations as they work towards their social and economic development goals (Barder 2012 Ramalingam 2013 Woolcock 2014) Nonetheless the dynamics of WASH service systems in low- and middle-income countries cannot be understood without considering the role that international development aid plays in setting national priorities The literature about the role and influence of international development aid in shaping national development agendas questions the planning horizons maintained by development partners do they allow sufficient time to achieve lasting systemic change (Rogers and Hall 2003) The literature also indicates that policy finance and political-economic priorities cannot be assumed to be aligned with national development agendas (Rogers and Hall 2003 Mowles et al 2008 Ramalingam 2013)

Level Frequency(Years)

Purpose

102 to 103

Often noncalculative

spontaneous (caveat see

discussion in text)

Get the institutional environment right

1st order economizing

Get the govemance

structures right 2nd order

economizing

Get the marginal

conditions right3rd order

economizing

10 to 102

1 to 10

continuous

L1

L2

L3

L4

FIGURE 2 ECONOMICS OF INSTITUTIONS

Source Williamson 2000 p597

Embeddednessinformal

institutionscustoms

traditions norms religion

Institutionalenvironmentformal rules of the game-epsproperty (polity

judiciary bureaucracy)

Govemanceplay of the game-esp

contact(aligning

govemance stuctures with transactions)

Resourceallocation and employment(prices and quantities incentive

alignment)

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 13: Change in complex adaptive systems

12

WORKING PAPER Change in complex adaptive systems

23 CAPITAL INVESTMENTS BUT FAILING SERVICES

Despite significant investments by both governments and their development partners the rural water sector is far from achieving the goal of safe reliable sustainable service for everyone everywheremdashand in particular the very poor This section reflects on the challenges faced by nations striving to achieve the goal of universal water services

In 2012 lsquoan estimated 22 of the worldrsquos rural population (740 million people) [did] not access a safe drinking water supplyrsquo (RWSN 2012 p7) The scope of the problem is vast lsquomore than 600 million of the estimated 700 million people who lack access to improved water services live in rural areasrsquo (Schouten and Moriarty 2013 p7) In 2010 lsquofive out of six people without access to a safe drinking water supply reside in rural areasrsquo (UNICEFWHO 2010) And despite trends towards urbanisation the number of rural dwellers will still be about 29 billion by 2050 (UNPD 2009) with the highest concentration of rural dwellers in Africa and Asiamdashregions that face the greatest challenges in advancing human and national development agendas

Following decades of prioritising the construction of new water infrastructure it is now widely recognised that new construction alone will not solve the problem WaterAid Tanzania reported that only two years following installation 25 of systems were already non-functional (Taylor 2009)

The multiple causes of the failure of the rural water sector are relatively well known (eg Lockwood and Smits 2011) Schouten and Moriarty (2013) list these inextricably interlinked causes bull Some national governments ignore the rural water

supply sector capital investment comes largely from development partners

bull Interventions by development partners are often uncoordinated stand-alone projects each with its own design hardware type policies and financingmdashprecluding efficiencies and coordination

bull The usual approach to rural water supply servicesmdashvillage-level operations and maintenance demand-response community managementmdashassumes that users can sustain service delivery without outside help

bull National water sectors often lack the vision strategy and capacity to sustain services

bull Lack of long-term planning for rural service delivery results in irregular unreliable supply

bull Financial models for sustainable service delivery and eventual replacement of infrastructure are missing leading to ad hoc provision of services

bull Systems fail before the design lifetime wasting capital sometimes multiple reinvestments are made in the same communities

Clearly there is no single or linear solution that can or will resolve these interlinked challenges and increase levels of access to water services As discussed in Section 21 the actors organisations formal and informal institutions (norms values policies shared strategiesmdashafter Ostrom 2011) involved in service development and delivery in a multi-level polycentric entity or system have overlapping areas of responsibility A business-as-usual approach to developmentmdashmaking linear uncoordinated interventions in an attempt to build resilient national systems that can deliver lasting servicesmdashis not working (Ramalingam 2013 Mowles et al 2008 Rogers and Hall 2003)

A range of approaches have emerged that seek to foster systemic change by engaging the whole system of actors and institutions involved in the delivery of common public goods These are discussed in more detail in Section 4 First however Section 24 reviews literature on the need to re-think international development aid to gain deeper insight into the challenges to the current aid approaches and to identify alternatives from the perspective of experts in international development aid

24 CALLS TO RE-THINK INTERNATIONAL DEVELOPMENT AID

Many stakeholders have called for change in how international development aid is conceptualised and implemented to improve performance of the water service and other sectors Nobel laureate Amartya Sen described the aim of social and economic development as lsquoenlarging peoplersquos choices capabilities and freedoms so that they can live a long and healthy life have access to key knowledge a decent standard of living and participate in the life of their communityrsquo (Sen 1992 cited in Barder 2012)

This human development perspective is also embodied in the United Nations Development Programmersquos Human Development Reports Indeed the 2014 report was entitled Sustaining Human Progress Reducing Vulnerabilities and Building Resilience Another justification for aid is the economic growth perspective expressed in traditional economic measurements such as gross domestic product The sustainable development perspective arose from works in the 1970s consolidated in 1987 in a United Nations World

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 14: Change in complex adaptive systems

13

December 2015

Commission on Environment and Development report Our Common Future (lsquothe Brundtland Reportrsquo) This perspective has since evolved through the Rio conventions of 1992 and 2012 which developed the Millennium Development Goals and at time of writing the draft post-2015 Sustainable Development Goals

Regardless of the measure one prefers enlarging peoplersquos choices capabilities and freedoms requires the accessible affordable provision of basic public goods and services that contribute to peoplersquos well-being Providing such services requires being able to act in an agile adaptive manner in the face of rapid socio-economic change and future uncertainty about climate change social stability and economic pressures (Barder 2012 Mowles et al 2008)

How then can the failures of rural water and sanitation services be addressed

Calls for a paradigm shift in the way development aid and interventions are conceptualised organised and function come from Barder (2012) Pritchett et al (2013b) Andrews et al (2012) Kania and Kramer (2013) and Woolcock (2014) These researchers discuss theory and practice that are grounded in complex adaptive systems thinking

Andrews et al (2012) propose a lsquoproblem-driven iterative approachrsquo (Section 4 below) and advocate a departure from linear simplistic approaches to implementing lsquosolutionsrsquo in favour of local processes that address specific problems by identifying and testing alternatives (Andrews et al 2013 Woolcock 2014)

Barder (2012) finds that complexity theory has implications for development policy He borrows from Senrsquos capabilities perspective and defines development as the lsquoemergence of a system of economic financial legal social and political institutions firms products and technologies which together provide citizens with the means to live happy healthy and productive livesrsquo (Barder 2012) The non-linear dynamics of such a system Barder believes can produce startling changes as agents within the system as well as the system itself adapt and co-evolve in response to one another He suggests the inevitability of lsquospontaneous rapid change to a more complex self-organised system which does a better job of supporting the capabilities of their citizensrsquo (Barder 2012) For these reasons Barder argues that the instrumental linear view of development should be abandoned in favour of policy

and implementation practices that enable actors to anticipate and adapt to unforeseen changes

Ramalingam argues for transformation in how the development aid system works starting from the level of lsquothe ldquorules of the gamerdquo that shape what can and canrsquot be done in aid that shape behaviours and actions that determine rewards and punishmentsrsquo (Ramalingam 2013 p16) Examples cited by Ramalingam illustrate how the development aid system and its problems are interconnected diverse and dynamic spanning layers of social institutional and political economies in different settings The aid system is a lsquomany to manyrsquo world with lsquomore agencies using more money and more frameworks to deliver more projects in more countries with more partners employing more staff specializing in more disciplinesrsquo (Ramalingam 2013 p5) But rather than calling this hyper-inter-connectedness a problem Ramalingam seeks to show lsquohow the ideas of complex systems research have been used to make aid ideas and aid practices more sensitive to the real-world dynamics of social economic and political phenomenarsquo (Ramalingam 2013 p244) He supports a transformation in lsquothe fundamental assumptions ideas and actions of aidrsquo based on the following observations (Ramalingam 2013 p360)bull the common mismatch between aid and the

challenges it strives to addressbull the imperfect and ambiguous nature of the effect of

aid bull the importance of domestic institutions and

political economy bull the dynamic nature of political transformations and

their context and bull the increasingly rapid pace at which change is

taking place

In this section a range of views articulated by domain experts and leading thinkers from the field of international development has been presented The literature cited here not only supports the finding in Section 22mdashthat prevailing international aid practices are misaligned with national development agendasmdashbut also underscores the value of a complexity-informed approach by development partners It has also shown the need to delve further into the concepts and theories of the complexity sciences to obtain a more complete and meaningful analysis of the rural water sector and its dynamics as a system

Section 3 explores how change in a complex adaptive system occurs and presents the central concepts and theories from the complexity sciences

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 15: Change in complex adaptive systems

14

WORKING PAPER Change in complex adaptive systems

In Section 3 concepts from literature about complex adaptive systemsmdashand in particular one sub-type of system a socio-technical systemmdashare presented Importantly this section focuses on ways to understand how change arises in such systems and reviews the literature on whole system change Previous sections have established the domain of inquiry and challenges faced by nations in delivering sustainable water services including challenges posed by prevailing policy and practices in international development aid Based upon the literature reviewed the case is made that prevailing arrangements for development and delivery of sustainable public services as well as for international development aid would benefit from the adoption of complexity-informed policies and practices This section introduces concepts and theories from the complexity sciences and then in Section 37 frames the rural water sector from a complexity perspective to gain insights into how and under what conditions systemic change might occur

31 COMPLEX ADAPTIVE SYSTEMS AND THE SOCIO-TECHNICAL SYSTEM

A complex adaptive system (CAS) is a dynamic network of many agents (whether cells species individuals firms or nations) acting in parallel constantly acting and reacting to what the other agents are doing The control of a CAS tends to be highly dispersed and decentralized (Ryan 2008) If there is to be any coherent behaviour in the system it has to arise from competition and cooperation among the agents themselves The overall behaviour of the system is the result of a huge number of decisions made every moment by many individual agents (Waldrop 1992)

A notable type of CAS useful for framing the rural water sector from a complexity perspective is the socio-technical system Socio-technical systems comprise lsquotwo deeply interconnected subsystems a social network of actors and a physical network of technical artefactsrsquo (Dijkema et al 2013 p1) These systems consist of lsquoheterogeneous decision making entities and technological artefactsrsquo and lsquoare governed by public policy in a multi-scale institutional contextrsquo (Ghorbani 2013 p3)

As the review of literature on the governance perspective in Section 21 indicates the delivery of public services in such a context requires interaction

among many diverse actors The provision of services such as energy solid waste removal water for domestic or commercial use and hygiene and sanitation services involves continuous and evolving interactions between the socio-political technical financial environmental and institutional realms

hellip[s]ocio-[t]echnical [s]ystems are [a] class of systems that span technical artefacts embedded in a social networkhellip[and] include social elements such as operating companies investors local and national governments regional development agencies non-governmental organizations customers and institutions These develop around sustain and depend on particular technical systems be it a single plant industrial complex or set of interconnected supply-chains (Nikolic and Ghorbani 2011 p1)

Because the WASH sector involves intertwined technical and social systems it fits the following definition of CAS

[A] multi-actor network determines the development operation and management of the technical network which in turn affects the behaviour of the actors The interactions within and between technical systems are defined by causal relationships which are governed by laws of nature while the actors in the social system develop intentional relationships to accomplish their individual goals At multiple hierarchical levels the technical network is shaped by the social network and vice-versa with feedback loops running across multiple levels and time scales All of this together forms a self-organising hierarchical open system with a multi-actor multi-level and multi-objective character (Holland 1992 cited in Dijkema et al 2013 p2)

This understanding of socio-technical systems enables further exploration of how processes concerning water service delivery across a multi-scale institutional context change over time A CAS perspective makes it possible to identify the macro-level emergent change patterns that arise from micro-level decision-making processes and interactions within a socio-technical system

In considering the potential for systemic change in the rural water sector as well as how best to foster this change a complex adaptive systems perspective is applied lsquoto stimulate and support the development of more flexible more reliable and more intelligent infrastructures and services with respect for public

3 Concepts from the complexity sciences

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 16: Change in complex adaptive systems

15

December 2015

values and consumer interest to better serve society in the futurersquo (Dijkema et al 2013 p7) The application of a complex adaptive systems perspective connects the literature of this domain with the governance perspective as discussed in Section 21 The next section presents a review of literature about how change arises in such systems

32 CHANGE IN SOCIO-TECHNICAL SYSTEMS

The review of literature about the governance and complexity perspectives suggests that change in such systems can be fostered This section therefore presents an overview of literature about the notion of whole system change as a phenomenon that can occur and has occurred in the domains of public services and development aid It reviews studies that propose ways to understand and speak about how large socio-technical systems evolve and adapt The remainder of Section 3 provides different perspectives on how such change happens what is actually changing when a system changes

The academic literature on the concept of whole system change is relatively modest but over the past two decades the concept has been increasingly featured in health care reform in the United Kingdom and Canada (Connor and Kissen 2010 Edwards et al 2011) educational reform in the United States (Duffy et al 2006) integrated water resources management (Pahl-Wostl et al 2007 Pahl-Wostl et al 2013) and to a limited extent in the development aid sector

Harman (1995 p1) has examined the plausibility of lsquowhole-system changersquo in the face of what he called lsquoglobal dilemmasrsquo such as anthropogenic climate change chronic hunger environmental degradation and poverty he proposes that these dilemmas are lsquonot so much problems as symptoms of a deeper-level condition that must be dealt withrsquo

Harman is not alone in proposing that nothing short of whole system change can address wicked problems Bramson and Buss (2002) published an overview of methods for whole system change in public organisations and communities Their work refers to lsquolarge group methodologiesrsquo as processes that involve lsquothe whole system both internal and external stakeholders in the change processrsquo (Bramson and Buss 2002 p212) Some of the large-group methods referenced in the literature include future search appreciative inquiry Whole-ScaleTM Change Participatory Strategic Planning Process Real Time Strategic Change and SimuReal

Large-group change methods are historically intertwined One strand emerged from theory about systems and how this has shaped modern views on organisations the second strand involves the technology for working with large systems and channelling the energy of a group into lsquoplanning for the future rather than focusing on problems and involving as much of the systemrsquo as possible to identify what works and aim for consensus (Bramson and Buss 2002 p214) The works reviewed by Bramson and Buss (2002) have several common elements the value system of democracy pluralism pragmatism activism self-expression and open communication as ways to lsquoovercome unnecessary obstacles to consensus and collective action among people with diverse interestsrsquo (Bramson and Buss 2002 p215)

Bramson and Buss (2002) also identified the following seven characteristics of whole system changebull Future driven Proponents assume that a shared

vision enables people to move past conflict and motivates them to action

bull Broadly participative Large numbers of people (hundreds thousands) from an organisation or community are engaged in understanding the interconnections among organisations interests or relationships This shared cognition enables them to participate and help make important decisions

bull Planning intensive Planning features in each of the methods reviewed and is considered the key to fostering stakeholdersrsquo buy-in

bull Skilled process facilitation Although Bramson and Buss (2002) mention this characteristic as a sub-element it is listed here in its own right to emphasize its importance in ensuring consistent and cohesive design and facilitation of a change process built on coalitions and human competencies An individual or organisation is required to curate and nurture the change process

bull Information sharing Commitment to sharing information with the whole system of people and organisations is founded on the belief that the people in the system have the lsquowisdom to know what is bestrsquo (Bramson and Buss 2002 p216)

bull Appeal to head and heart The methods reviewed appeal to both the intellect and the ethos of the people within a system so that they see the whole system and play a meaningful role in making things happen

bull Sustainability A series of connected events with coordinated and coherent agendas happening at agreed intervals with each event possibly spanning multiple days is crucial for the change process which requires dedicated ongoing facilitation as

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 17: Change in complex adaptive systems

16

WORKING PAPER Change in complex adaptive systems

well as follow-up by stakeholders on implementing the agreed strategies and action plans in their own organisations

Although other criteria may yet be required for a complete understanding of what such an approach entails those seven characteristics form an initial series of elements that can be used to formalise and compare different approaches to effecting whole system change

Bramson and Buss anticipate that whole system change approaches will proliferate because of pressure on development organisations to produce desirable results quickly the availability of facilitators experienced with engaging large groups in systemic change processes and an increase in familiarity with the approaches in different sectors They also point to wider acceptance of the idea that change in a world of interconnected systems is best understood through systems thinking informed by lsquovarious parts of the relevant systemrsquo (Bramson and Buss 2002 p218) in the same room and that democratisationmdashfrequently supported by international development aidmdashassumes that lsquobetter decisions hellip result from involving more people in public decision-makingrsquo ( Huitema et al 2009)

According to Burns (2007) decision makers need to provide more space for solutions to emerge from inquiry and learning processes as opposed to deciding in advance what a solution is testing it and rolling out the same model in other contexts (Burns 2007 p174 178) Similarly Brinkerhoff (2010) urges policy makers and implementers to adopt a systems perspective that favours incremental and emergent approaches to policy change that are informed by the voices of the excluded result from shared inquiry and dialogue and promote open and transparent decision making and citizen empowerment

Brinkerhoff summarises the seven design principles that Burns considers necessary for systemic action research lsquoemergent and flexible research design exploratory inquiry phase multiple inquiry streams at different levels connecting inquiry to formal decision making process to identify links across inquiry streams recognition that inquiry stream membership changes over time and commitment to distributed leadershiprsquo (Brinkerhoff 2010 p94) This set of design principles is useful for elaborating on Bramson and Bussrsquos seven characteristics because they introduce the foundations of learning evolution and adaptation of complex adaptive systems

The literature also provides case studies that support the possibility of whole system change in their depiction of strategies approaches and methods for implementation (White 2000 Manning and De la Cerda 2003 Dattee and Barlow 2010) Duffy et al (2006) provide a protocol for whole system change in school districts Their iterative process consists of a pre-launch preparation phase and three steps followed by a recycle to the next pre-launch preparation phase This seemingly simplified protocol belies their observation that lsquoa significant change in one part of the school system requires changes in the other parts of the systemrsquo (Duffy et al 2006 p41)

In the context of integrated water resources management Pahl-Wostl et al (2007 2013) focus on lsquotransformative changersquo arising from multi-level social or lsquosocietalrsquo learning and adaptive management approaches for achieving paradigm changes where system elements such as actors organisations infrastructure knowledge and power relations are highly interdependent Pahl-Wostl (2009 p354) developed a conceptual framework for use in analysing adaptive capacity and multi-level learning processes to enable deeper insights into lsquocomplex and diverse resource governance regimesrsquo

Greenhalgh et al (2012 p516) describe a lsquotransferable methodology developed to guide the evaluation of a three-year follow up of a large health care change programmersquo that took place in London during lsquoa period of economic turbulence and rapid policy changersquo This work gives attention to the tension that arises in large-scale change because of the persistence of past practice and the need to adapt to a changing context Tracking what lsquosurvivedrsquo three years after modernisation of a large health service Greenhalgh et al (2012) derived five conclusions about approaches to fostering whole system change bull To assess the effect of a large-scale change on

turbulent and dynamic settings one needs to ask not only lsquowhat has remainedrsquo from the originally intended programme outcomes but also lsquohow have things moved on and whyrsquo (p540)

bull A whole system change perspective is critical to ensuring that programme activities and outcomes succeed in lsquo[l]inking the transformation effort more closely to the mainstream-commissioning and business-planning infrastructurersquo despite the potential for this to slow the rate of change (p540)

bull lsquo[T]he knowledge hellip to sustain complex service innovations spanning multiple organizations and sectors appear[s] to be largely tied to individuals embedded in relationships and strongly value ladenrsquo (p540) Relationships that are lsquowarmrsquo strengthen

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 18: Change in complex adaptive systems

17

December 2015

shared priority-setting and participants identify solutions more rapidly in response to dynamic and changing circumstances

bull lsquo[T]ransferable modelsrsquo may not be realistic given the need to continually adapt interventions lsquoin real time as the program takes shapersquo (p541)

bull A series of questions can prompt a shift in focus from lsquologic modelsrsquo or established and possibly rigid ways of framing an issue towards individual and group priorities for the allocation of resources emerging points of convergence and divergence and alignment of the programme with stakeholdersrsquo priorities lsquoin a tight quality cyclersquo (p541)

The authors recognise that because many stake holders may not be familiar with whole system change its success lsquodepends upon achieving widespread confidence and capability to go beyond logic-modelsrsquo that are linear and control-oriented in framing issues or challenges (Greenhalg et al 2012 p541)

The case studies indicate that the process of whole system change is neither linear nor simple The clicheacutes apply there are no panaceas silver bullets or quick fixes to address the interconnected failings across a socio-technical system whose problems have evolved over time and largely become intractable Literature from both the governance perspective and the complexity complexity-informed perspective addresses the need for identifying the context-specific nature of challenges along with locally relevant solutions that receive popular understanding and support (Huitema et al 2009 Mowles et al 2008 Burns 2007 Bramson and Buss 2002)

What is common among these methods is perhaps best summarised by the categories set out by Huitema et al (2009) approaches to effecting change that recognise the polycentric nature of public services involve public participation employ experimentation and are bio regional in nature In their work on water resources manage ment lsquobioregionalrsquo refers to river basins as the relevant scale at which to conceptualise the system under examination (Huitema et al 2009 p9) In essence their bioregional approach echoes the locally relevant analysis and solution identification described by other researchers

The literature establishes whole system change as a concept Several researchers also attempt to distil its essential characteristics principles or elements that may be applied in analysing and formalising such approaches Nonetheless unresolved questions include whether certain approaches are more effective than other approaches and whether upon inspection

and comparison specific elements are more effective than others The following section therefore delves into additional concepts and theory from the complexity sciences that offer both a meta-theory of how systems evolve as well as a grammar about how institutions change That allows us to describe and analyse the formal and informal rules and shared strategies in human behaviour that guide the micro-level actions and interactions that give rise to overall patterns and trends in a given system

33 UNIVERSAL DARWINISM A META-THEORY OF EVOLVING SYSTEMS

lsquoOne general law leading to advancement of all organic being namely multiply vary let the strongest live and the weakest diersquo (Darwin 1859)

Commonly referred to as universal Darwinism the body of theory introduced below provides lsquoa general or meta-theoretical frameworkrsquo (Hodgson 2008 p404) to thinking systematically about processes of emergence and change in complex social and institutional systems (Aldrich et al 2008 Hodgson 2008 Hodgson and Knudsen 2010) Universal Darwinism suggests that institutions information and organisations emerge and evolve in ways similar to those laid out by Darwin through selection variation and retention (Blyth et al 2011)

In 1898 an American economist and social scientist asked Why is economics not an evolutionary science (Veblen 1898) He articulated the view that Darwinrsquos theory of evolution and its associated processes of selection variation and heredity were relevant to understanding social institutions and how these structures emerge and change over time (Hodgson 2008 p44) He considered evolutionary science a lsquoclose-knit body of theoryrsquo that could reliably explain the evolution of social as well as biological phenomena (Veblen 1898 p404)

Veblen proposed that social evolution was a natural selection of institutions

The life of man in society just like the life of other species is a struggle for existence and therefore it is a process of selective adaptation The evolution of social structure has been a process of natural selection of institutions The progress which has been and is being made in human institutions and in human character may be set down broadly to a natural selection of the fittest habits of thought and to a process of enforced adaptation of individuals to an environment which has progressively changed with the growth of community

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 19: Change in complex adaptive systems

18

WORKING PAPER Change in complex adaptive systems

and with the changing institutions under which men have lived (Veblen 1899 p188)

Modern scholars have acknowledged Veblenrsquos theoretical contribution while adapting it to examine the fields of political science (Lewis and Steinmo 2010) institutional evolution and change (Lewis and Steinmo 2012) learning selection in uptake and adaptation of new technologies (Douthwaite et al 2002) organi-sational learning and change (Trist 1981 Stoelhorst and Huizing 2006) and human language and business corporations (Hodgson and Knudsen 2010)

Universal Darwinism enables the following questions to be asked How do social phenomena act or interact What patterns of behaviour habits or beliefs are common Which actions or interactions adapt evolve and carry on Do certain behaviours practices habits or beliefs die out The premise is that social pheno-mena are subject to Darwinian processes of evolution as accepted in the natural sciences However as Hodgson (2008) notes additional theories and tools are required to more precisely describe social phenomena the patterns they exhibit and the pressures that effect change in a particular context as well as to gain insight into the broader line of inquiry of how to foster change within a whole system such as the rural water sector For this reason the next section discusses a framework for conceptualising analysing and structuring institutional change as proposed by Elinor Ostrom in her institutional analysis and develop ment framework (Ostrom 2011) The frame-work offers a grammar for making explicit the formal and informal rules or institutions that shape human behaviour and patterns of interaction

34 THE LANGUAGE OF INSTITUTIONS IAD AND ADICO

This section introduces two frameworks that can serve as the basis for the structured identification and analysis of formal and informal institutions Both were devised by the late Nobel laureate Elinor Ostrom who sought to explain how institutional rules norms and strategies change over time (Ostrom 2011) Institutions are a set of devised rules to organise repetitive activities and shape human interaction (Gardner and Ostrom 1991 Ostrom 2011) Understanding institutions in this manner makes it possible to lsquoconceptualise the dynamic interplay between actors and structuresrsquo (Geels 2004 p897)

In her doctoral thesis Ghorbani describes the two frameworks of institutional analysisbull ADICOmdashattributes (participants) deontic (obligated

permitted forbidden etc) aim (action or outcome) condition (parameters when an ADICO statement applies) or else (sanction) (Crawford and Ostrom 1995) This is the lsquogrammarrsquo of institutions (Crawford and Ostrom 1995)

bull IADmdashinstitutional analysis and development (Polski and Ostrom 1999)

These frameworks offer structure to the analysis of socio-technical systems by recognising that individual behaviour is complex and not easily extractable whereas formal social rules and institutions are fairly well extractable and therefore measurable (Crawford and Ostrom 1995 Ghorbani 2013)

Institutional grammar is useful for examining the different foundations (norms laws and shared strategies) of different types of institutions (Crawford and Ostrom 1995) The grammar defines laws as ADICO norms as ADIC and shared strategies as AIC That is unlike shared strategies and norms laws alone have sanctions (the o from lsquoor elsersquo)

Crawford and Ostrom establish the grounds for considering institutions as norms and regulations whereby institutions as norms assumes that many patterns of interaction are based on a group of individualsrsquo shared perceptions about proper and improper behaviour in particular situations (Crawford and Ostrom 1995) and institutions as regulations assumes that many patterns of interaction are based on a common understanding Forbidden or proscribed actions are likely to be sanctioned or made ineffective (if an authority imposes punishment) (Crawford and Ostrom 1995)

These frameworks offer a means of understanding why certain regularities of human behaviour exist It is important to note an institutionrsquos explanation for behavioural patterns and locate the responsibility for social order with the individuals who are part of that system not to some external state or third-party enforcer (Crawford and Ostrom 1995) This view integrates the analysis of how institutions come into place with the analysis from within In this manner an institution may be seen as a pattern of behaviour sustained by mutual expectations of the behaviour of others

Figure 3 separates the operational level from the institutional level On the institutional level the formal institutions and informal institutions are apparent The operational level shows two boxes priorities of the agents and decision-making process Also visible on this operational level are the

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 20: Change in complex adaptive systems

19

December 2015

institutions that are imposed on the humans or organizations that are making decisions Priorities of the agents are also taken into account in decision making (Polski and Ostrom 1999) and may be rational or irrational in other words agents are conscious and capable of self-reflection The decision may or may not be in line with the institution and if not consequences may ensue For example a driver who runs a red light places a higher priority on getting to his destination than on obeying traffic laws but risks causing an accident or getting a traffic ticket

Using the theoretical framework and language of institutions helps us identify and analyse the institutions that guide micro-level behaviours and decisions of agents in a given environment The ADICO and IAD frameworks offer a means of making explicit regularities of human behaviour and in combination with universal Darwinism the language of institutions helps explain how institutions emerge and evolve over time

Sections 31 through 34 have introduced concepts and theories from the complexity sciences about what constitutes a specific type of complex adaptive systemmdashthe socio-technical system The literature is explicit about the fact that whole system change in such systems can occur and several sources call for a whole system approach Additional concepts relate to how institutional change emerges and changes over time This body of theory makes it possible to test

theories about how the institutions or lsquorules of the gamersquo that guide how the social and technical artefacts act interact and may change Testing is an important step for furthering the line of inquiry on how to effect change in such systems

Section 35 now reconsiders the rural water sector from the perspective of a socio-technical system

35 RURAL WATER SERVICES AS A SOCIO-TECHNICAL SYSTEM

A system of interconnected social actors organisations and institutions (policies norms and beliefs) is required to ensure that water delivery infrastructure operates as intended over the lifetime of the technical system But is the rural water services sector in fact a complex adaptive socio-technical system This section examines the phenomenon of the rural water services sector in light of CAS and socio-technical system theory

Drawing on Dijkema et al (2013 see Section 31 above) Van Dam et al (2013) characterize complex adaptive systems as followsbull Multi-actor many different (heterogeneous) actors

or agents act and interact with intention through social networks

bull Multi-objective different actors within the system hold different priorities

Institutional level

Laws Norms

Formal institutions

Sharedstrategies

Informal institutions

Operational level

Priorities ofthe agents

Decisionmakingprocess

Decision

Outcome of decision making process which is in line with

the institutions or not

Consequence for the decisionmaker and system

A complementary view Institutional analys is for describing socio-technical systems

An institution is a ruleof behavior imposed on you

by society

ADDITION

FIGURE 3 IAD FRAMEWORK Source Adapted from Polski and Ostrom 1999

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 21: Change in complex adaptive systems

20

WORKING PAPER Change in complex adaptive systems

bull Feedback loops connections (eg information or financial flows) run across the hierarchical levels time scales individuals and social networks (Dijkema et al 2013 p2)

bull The system is self-organising hierarchical and open (SOHO) a Self-organising lsquothe process by which a system

develops a structure or pattern without the imposition of structure from a central or outside authority or when a system displays a different output as a result of internal processesrsquo (Prigogine and Stengers 1984 Kay 2002 cited in Nikolic and Kasmire 2013 p50)

b Hierarchical multiple hierarchical levels and c Open lsquowhere matter and energy [and

information] flow in and out and where things inside the system are affected by the environ-ment outside the systemhelliprsquo (Nikolic and Kasmire 2013 p15)

As an example Figure 4 depicts the landscape of actors and institutions in the rural water service delivery sector in the Republic of Uganda a vast constellation of interconnected agents across multiple levels (community district municipal regional national) all with the shared aim of ensuring access to water services by rural populations

From this image several features of a complex adaptive system are immediately observable The rural water services sector comprises multiple actors or agents who act and interact through social networks These actors hold different but occasionally overlapping priorities Each agentmdashindividual organisation or networkmdashhas capabilities beliefs values skills and resources that evolve over time and guide how it acts and interacts with others This high degree of interconnectedness creates multiple feedback and feed-forward loops actions interactions and networks are interconnected through information financial and human resources trust directives etc spanning hierarchical levels and time scales (Dijkema et al 2013 p 2) The multi-stakeholder platforms depicted in Figure 4 play a key role in shaping sector priorities and thus this water sector is self-organising Lastly the system receives and provides flows of information financial and other resourcesmdashinternational development aid water from the ecosystem macro-level economic systems etcmdashensuring that the system is open and may be shaped or affected by the lsquoenvironment outside the systemrsquo

It is not possible to depict the dynamic and evolving nature of Ugandarsquos rural water sector in a two-dimensional image Nevertheless Figure 4 provides a

starting point to understand what is inside versus outside this socio-technical system Such an image can also help identify bottlenecks to innovation change or improved services as well as weak points in the interconnections and feedback and feed-forward loops

Following this framing of the rural water sector as a complex adaptive socio-technical system Section 4 presents three specific approaches and their guiding tenets principles and strategic objectives for whole system change These three approaches are generally considered well suited for fostering systemic change in the context of national and international development initiatives and public service delivery processes

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 22: Change in complex adaptive systems

21

December 2015

FIGURE 4 REPUBLIC OF UGANDA RURAL WATER SECTOR ACTOR LANDSCAPE Source Casella et al (2013)

JSR-SH Forum

MWEDWD+ all other arms

NLF

JTR

DPrsquoS

OthersMIS

functionality

UWASNETCSO Wash Forum

TSU review

IDMrsquoS

(bi-)annualRegional Learning

DWSC

MWEDWD

Privatesector

Media

AcademiaDWOrsquos

CSOrsquos

Political amp Technicalleading

TSU26

Regional

DWOrsquos

MWEDWD

Cao or LCS

Privatesector NGOrsquos

DWO

LCV

CSOrsquos

Media

TSU

NWSC

HPMA Chair

Distr Healt Dept

Sub-CountylsquoSWSCCrsquo

District

Sub County

ParishesVillages

Chair-S-C Chief

Local Country 3

2 HPMrsquos

Secretaryassistant CDO

Health Assistent

CSOrsquos(incl NGO+CBO) (Serv Prov)

Private operators

(serv providers)

NAADS Officer

Parish chiefs

Lira15 DWSCCrsquosrep by DWO

+ CSOrsquos

Kabarole10 DWSCCrsquosrep by DWO

Reports on DistrictWater status-compiled

From sc reports

-Health-Productions

Dw

d se

nds

plan

s ba

ck

to C

DO

rsquos

Non Mandetory

Mandetory

Info Flows

Reports

Repo

rts

go to

DW

O P

rovi

ded

by

asst

CD

OS

ECT

CAO

rsquos C

SOrsquos

amp H

PMrsquos

CAO

Plan Devine ILF Caritas LTP SNV

NoWashH person Specialistfunction

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 23: Change in complex adaptive systems

22

WORKING PAPER Change in complex adaptive systems

In Section 3 theory and concepts from the field of complexity sciences enabled the framing of the rural water sector as a complex adaptive system The notion that change in such systems is possible and that whole system change approaches are suited for fostering such change was discussed In this section three approaches to fostering large-scale change in complex systems are described collective impact problem driven iterative analysis and learning alliances These approaches have been documented in various papers as having relevance to fostering change promoting innovation and scaling up promising solutions in different public service sectors

41 COLLECTIVE IMPACT CREATING LARGE-SCALE SOCIAL CHANGE

The first approach to effecting social change is called collective impact a term coined by Kania and Kramer (2011) following a decade of experience and research on effective social change processes The approach is based on their observations of public education system reform in the United States They frame the case example and its analysis in the language and concepts from complexity sciences and chaos theory

Their 2011 case study details the experience of Strive a nonprofit to lsquobring together local leaders to tackle the student achievement crisis and improve education throughout greater Cincinnati [Ohio] and northern Kentuckyrsquo (Kania and Kramer 2011 p36) The crux of the case study is that those seeking to effect the change realized that lsquofixing one point on the educational continuum hellip wouldnrsquot make much difference unless all parts of the continuum improved at the same timersquo (p36) Strive therefore focused the educational community on one set of goals measured in the same way The Strive experience had lsquocollective impactrsquo because the initiative involved lsquoa centralized infrastructure a dedicated staff and a structured process that leads to a common agenda shared measurement continuous communication and mutually reinforcing activities among all participantsrsquo (p38) Other collective impact initiatives include watershed restoration along the Elizabeth River in Virginia a childhood obesity prevention programme in Somerville Massachusetts and agricultural sector reform for private and social benefits in Cote drsquoIvoire The unifying element of these examples is that large-scale social change lsquocomes from better cross-sector coordination rather than from the isolated intervention of individual organisationsrsquo (p38)

In earlier work Kania and Kramer (2004) distinguished between adaptive and technical problems In the latter the problem and its solution are commonly known and change can be made by one or a few organisations An example of a social problem that they define as lsquotechnicalrsquo is running a scholarship programme In contrast lsquoadaptive problems hellip are complex the answer is not known and even if it were no single entity has the resources or authority to bring about the necessary changersquo (Kania and Kramer 2011 p39)

Moving from isolated impact where success in one location cannot be replicated or scaled in another setting to collective impact is lsquonot merely a matter of encouraging more collaboration or public-private partnershipsrsquo

[Collective impact] requires a systemic approach to social impact that focuses on the relationships between organizations and the progress toward shared objectives And it requires the hellip organizations that have the skills and resources to assemble and coordinate the specific elements necessary for collective action to succeed (Kania and Kramer 2011 p39)

That is the lsquoprocess and results of collective impact are emergent rather than predetermined the necessary resources and innovations often already exist but have not yet been recognised learning is continuous and adoption happens simultaneously among many different organisationsrsquo (Kania and Kramer 2013 p2) Based on additional examples of the structured approach the authors identify the following five conditions for collective impact (Kania and Kramer 2011)bull a common agenda or shared vision for change

among all participants based on a common understanding of the problem and a joint approach to solving it

bull a shared measurement system that captures data in a short list of indicators that are agreed and reported by all participants as the measures of success

bull mutually reinforcing activities that recognize the interconnectedness across organisations actors and outcomes and are coordinated even though they may be highly diverse

bull continual communication that builds trust and creates a common vocabulary as the basis for the shared measurement system including face-to-face communication newsletters minutes and other records of meetings and public fora and

4 Approaches to whole system change

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 24: Change in complex adaptive systems

23

December 2015

exchange of information and experiences among participants and

bull backbone support organisations staffed by dedicated people with highly specific skills who can manage drive and coordinate the previous four criteria

Although those criteria seem to focus on the process required to achieve change as Kania and Kramer observe when emergent solutions begin to meet the intentional outcomes lsquothe process becomes the solutionrsquo (Kania and Kramer 2013 p7)

There are also challenges to creating lasting change through collective impact whereby capturing learning does not necessarily lead to acting on the lessons learned adequacy (or lack thereof) of resources such as money time and capacity and the difficulty of securing funding for lsquointentional change with emergent solutionsrsquo approaches in current development and public services paradigms (Kania and Kramer 2011 2013)

A review of case study examples of collective impact approaches was not conducted as part of this literature review Several such cases have been documented however and method reviews variously extol or critique the collective impact approach This line of inquiry will be researched in more detail in subsequent steps in this ongoing research project as the research progresses towards analysis and comparison of the approaches to fostering systemic change for insights into their relative similarities and differences

42 LEARNING ALLIANCES

The second approach to fostering large-scale socio-technical system change involves learning alliances In the WASH sector learning alliances were identified in

the early 2000s as a promising approach for scaling up innovations The term learning alliance has however been in widespread use in the business world since the end of 1980s (see Iyer 2002 and Khanna et al 1998) In other areas of development especially in agro-enter-prise development learning alliances were used in the 1990s so that people could come together to analyse problems address the challenge of mismatched expectations and interests and seek solutions The term is also used in health and educationmdashfor example in the National Health Service in the United Kingdom (Connor 2001)

The Colombia-based Centro Internacional de Agricultura Tropical defined learning alliances as follows

[a] process undertaken jointly by research organiza-tions donor and development agencies policy makers and the private sector through which good practices in both research and development are identified shared adapted and used to strengthen capacities improve practices generate and document development out comes identify future research needs and potential areas for collaboration and inform both public and private policy decisions (Lundy et al 2005 p3)

Centro Internacional de Agricultura Tropical follow this approach in the Rural Agro-enterprise Development Project (Lundy 2004 Lundy et al 2005) and advocate its use more widely in the Consultative Group on International Agricultural Research as a means of increasing the effectiveness and relevance of research the impact of development work and the formulation of better-informed policies

Learning alliances foster feedback loops that are both lsquohorizontalrsquo (among stakeholders working at the same

Vision Diagnosis

Years 1-2 Years 5-10Years 2-5

Solutions Testing Scaling up

build consensus build consensus build consensus build consensus

evidence refine evaluate adapt

FIGURE 5 SECTOR CHANGE AND INNOVATION PROCESS

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 25: Change in complex adaptive systems

24

WORKING PAPER Change in complex adaptive systems

institutional levelmdasheg civil servants and NGO staff) and lsquoverticalrsquo (eg between citizens and national government officials) across private public and non-governmental organisations academic and research organisations development partners civil society and other actors at different administrative levels This potential lsquoto bridge the gap between people on the ground organisations at district or provincial level with responsibility for service provision and support and national policy makersrsquo (Smits et al 2007 pxiii) is important for facilitating the scaling up and uptake of innovative solutions it engages actors in developing and testing solutions to problems (Figure 5)

The definition of a learning alliance used by IRC recog nises the layered structure of the WASH sector and refers to platforms at different institutional levels (Figure 6)

Moriarty et al (2005 p9) define a learning alliance as a series of interconnected lsquomulti-stakeholder platforms at key institutional levels (national district community etc) designed to break down barriers to both horizontal and vertical information sharing and thus to speed up the process of identification development and scaling up of innovationrsquo Da Silva Wells (2012) lists the strategic objectives for learning alliancesbull To provide dedicated space for deep reflection by

groups of stakeholders on specific WASH issuesbull To create feedback mechanisms between existing

multi-stakeholder platformsmdashboth horizontal (eg between districts) and vertical (eg between district and national levels)

bull To generate evidence on WASH challenges solutions innovations and opportunities to inform decision making and scaling up of proven and promising approaches to service delivery

bull To facilitate joint reflection analysis and action planning

bull To embed results of action research in the appropriate sector institutions or agencies

bull To create and support a critical mass of change agents

bull To accelerate the process of generating information joint reflection sense making planning and adaption through dedicated facilitation knowledge and information sharing

bull To influence policy processes

As with the principles of the collective impact approach those eight strategic objectives of learning alliances will serve in subsequent research steps as the basis for decomposition analysis and comparison of the three approaches discussed in this section

Learning alliances are a cornerstone of IRCrsquos approach to whole system change for sustainable water services Experiences in rural and urban settings across more than a dozen countries over the past decade have been documented by IRC and its partners in published case studies demonstrating the different forms and functions of learning alliances in a range of contexts The value and validity of the learning alliance approach have been assessed in various studies (Nkum et al 2014 Kahangire et al 2012) Approximately US$ 1 million per annum was required to support national and regional learning alliances in Ghana over six years (Duti and Lockwood 2015) Smits et al (2011) however find that few development partnersmdashdomestic or internationalmdashhave been willing to invest in making this lsquosoftwarersquo part of national WASH sectors

Given the availability of several case studies and documented examples of learning alliance

GLOBAL

NATIONAL

INTERMEDIATE

COMMUNITY

Scale up

Share lessons at a national level

Learn together

IntermediatePlatform

NationalPlatform

GlobalPlatform

Consolidate lessons Scale up Action reseach

Facilitator

FIGURE 6 LEARNING ALLIANCE APPROACH TO SCALING CHANGE ACROSS INSTITUTIONAL LEVELS

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 26: Change in complex adaptive systems

25

December 2015

approaches to delivering water services in different national contexts this approach is selected for analysis and comparison with collective impact discussed above and problem-driven iterative adaptation discussed below

43 PROBLEM-DRIVEN ITERATIVE ADAPTATION

Problem-driven iterative adaptation (PDIA) is the third complexity-based approach to fostering change and innovation in complex systems such as the delivery of public services

Andrews et al (2012) highlight the lsquocapability trapsrsquo that are the bane of institutional reform initiatives in developing countries In this and related work (Pritchett et al 2010 Pritchett et al 2013b Andrews et al 2013) researchers investigate the sources of lsquoimplementation failurersquo despite decades of institutional reform initiatives and attempts to scale up successful implementation practices from one context to another

Systemic isomorphic mimicrymdashthe lsquofake it till you make itrsquo strategy of adopting the appearance of another countryrsquos successful institution (see Section 22 of this review)mdashis highly problematic (Pritchett et al 2010) as an approach to economic and social development The desired resultmdashin the case of the WASH sector the sustainable delivery of WASH services to all citizensmdashcannot be achieved by simply creating attractive organograms of government entities when these entities lack human capacities financial resources reformed legal systems and enhanced public management systems (Pritchett et al 2013b) Andrews et al (2012 p7) therefore call on the international development sector to focus not on the form of organisations but their function and to shift towards lsquoa constant process through which agents make organizations better performers regardless of the form adopted to effect such changersquo

Following the review of literature ranging from Sengersquos (1990) work on lsquolearning organisationsrsquo through Grindlersquos (2004) lsquogood-enough governancersquo Briggsrsquos (2008) lsquodemocracy as problem solvingrsquo and Pritchett et alrsquos (2013a) lsquoexperiential learningrsquo Andrews et al (2012) propose the following four principles for problem-driven iterative adaptation for overcoming these problems of governance reformbull aim to solve particular problems in particular local

contexts via bull the creation of an lsquoauthorising environmentrsquo for

decision making that encourages experimentation and lsquopositive deviancersquo which gives rise to

bull active ongoing and experiential (and experimental) learning and the iterative feedback of lessons into new solutions doing so by

bull lsquoengaging broad sets of agents to ensure reforms are viable legitimate and relevantmdashthat is politically supportable and practically implementablersquo (Andrews et al 2012 p8)

Noting progress achieved through development assistance over the past decades Andrewrsquos colleague and fellow PDIA author Woolcock (2014 p24) advocates for further lsquoinstitutional changersquo and offers a more concise set of PDIA principlesbull local solutions for particular local problems bull pushing problem-driven positive deviance or a

lsquopurposive crawl of the design spacersquo as opposed to implementation of exhaustive plans made in advance

bull try learn iterate adapt following the lsquointegration of rigorous ldquoexperientialrdquo (and experimental) learning into tight feedback loopsrsquo and

bull scale up learning through diffusion lsquoof feasible practice across organizations and communities of practitionersrsquo (adapted from Andrews et al 2013)

Like collective impact and learning alliances PDIA is apparently a process-oriented approach However as noted explicitly by Kania and Kramer there comes a point at which the process steps start to merge with the intended outcomes of an intervention and the lsquoprocess becomes the solutionrsquo (Kania and Kramer 2013 p7)

PDIA recognises that interventions intended to achieve change have varying levels of complexity building a school is ldquoeasyrdquo lsquobuilding capabilities of the human systems hellip [is] hellip more difficultrsquo (Andrews et al 2013 p234) Knowledge about the context and the nature of a problem is fundamental as is the involvement of all stakeholders Robust legitimate public institutions are at the heart of the process of achieving change Functioning public institutions with the capacity to implement and adapt in the face of uncertainty is a critical aspect of overcoming the problems of the lsquocapability traprsquo and lsquoimplementation failurersquo (Pritchett et al 2013b p2)

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 27: Change in complex adaptive systems

26

WORKING PAPER Change in complex adaptive systems

The above three approaches feature in the literature about change and innovation in complex systems such as socio-technical systems for public services in low- and middle-income countries An analysis of the extent to which they let alone other approaches not reviewed here overlap or conflict was not found in the literature However for a policy maker or practitioner such insights would be invaluable for making decisions about how to allocate resources The literature is clear that development of new infrastructure alone will not achieve the goal of universal water services that last More resilient and adaptive national systemsmdashwhere the lsquoprocess becomes the solutionrsquomdashare required

To address this gap in the literature and knowledge base a framework for analysis and comparison of these approaches based upon Ostromrsquos institutional analysis and development framework (described in Section 3) will be developed The aim is to gain insights into the commonalities and differences across the approaches to whole system change An important line of inquiry is whether specific elementsmdashor even entire approachesmdashare more relevant or effective in fostering change

In this section tools for simulating complex problems and experimentation with alternative scenarios are presented These tools documented in the complexity sciences literature simulate the dynamics and emergent properties of complex adaptive systems including socio-technical systems and therefore offer a means for exploring the approaches reviewed in Section 4

Simulation is a powerful tool for learning and reflection when real-life testing and experimentation are not a realistic option (Ryan 2008 Flood 2010) This section reviews the available computational and non-computational modelling and simulation tools commonly used in the complexity sciences to support policy makers researchers practitioners and other stakeholders in exploring complex challenges Simulation of a problem offers opportunity for learning probing and testing policy or design assumptions or as Flood (2010 p277) suggests for

arriving at a lsquomeaningful understandingrsquo of the dynamics of a socio-technical system its actors and the prevailing institutions It offers a safe setting in which potentially promising solutions can be implementedmdashand their effects observed and analysedmdashwithout the risks or costs of testing in situ Using one or a combination of simulation methods in rigorous and structured approaches can provide insights into drivers and barriers to change in socio-technical systems

Two simulation tools well suited to exploring the effects of different interventions in the rural water sector through in silico8 experimentation are agent-based modelling and serious games (Ryan 2008 Ramalingam 2013) both of which generate macro-level outcomes arising from the individual behaviours decisions or actions of actors A review of the literature on agent-based modelling (Section 51) is followed by a review of the potential of serious games (Section 52)

51 AGENT-BASED MODELLING

This section presents literature that explains and discusses the value of agent-based modelling as a suitable tool for simulating and exploring complex challengesIn the 1960s Tom Schelling sketched maps of fictional racially segregated neighbourhoods to understand how lsquoindividual behavioural choices could aggregate into system-wide social phenomenon that were unintended unexpected and in this case undesirablersquo (Ramalingam 2013 p174) This application of social science to examine a social phenomenon as opposed to earlier applications in the physical sciences of genetics and biology was a first (Epstein et al 1996 Ramalingam 2013)

Agent-based modelling has since evolved from Schellingrsquos paper-and-pencil graphics to computer simulations of macro-level social patterns change or distributions that arise from micro-level processes actions and interactions in complex adaptive systems (Epstein et al 1996 Miller and Page 2007 Nikolic and Ghorbani 2011 Van Dam et al 2013) Agent-based models lsquoare constructed to discover possible

5 Simulating complex problems Tools to test and learn

8 ldquoin silicordquo [an action] done or produced by using computer software or simulation Merriam-Webstercom 2015 httpwwwmerriam-webstercom (13 May 2015)

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 28: Change in complex adaptive systems

27

December 2015

emergent properties from a bottom-up perspective They attempt to replicate in silico certain concepts actions relations or mechanisms that are proposed to exist in the real-world in order to see what happensrsquo (Nikolic and Kasmire 2013 p55) The modeller has no desired state or task to achieve instead the model lsquomerely describe[s] the entities and observe[s] how they interact in order to explore the systemrsquos possible statesrsquo (Nikolic and Kasmire 2013 pp55ndash56) Scientists policy makers and practitioners now widely lsquouse agent-based models to analyse socio-technical problems and explore policy alternativesrsquo (Ghorbani 2013 p3) in fields as diverse as energy markets public health care systems urban planning teamwork in organisations greenhouse agriculture systems supply-chain optimisation and financial markets (Van Dam et al 2013 Nikolic 2009)

Agent-based models are not predictive but they offer a means of exploring the arising dynamics in a defined problem space To arrive at a model that provides useful or meaningful insights a modeller begins by identifying the problem and then determines the agents rules states actions environment and flow of information as they are observed in the real world The simplified rules or institutions that are simulated guide how agents act and interact with one another and with the environment (Ghorbani 2013) Because of the bottom-up generative approach of such models the micro-level processes in turn

emerge as macro-scale patterns which may be non-linear (disproportional) to the initial micro-level processes as with the example of financial market booms Feedback loops may also be represented depicting the processes by which agents lsquolearnrsquo and adapt their behaviours over time Algorithms form the narrative of the simulated agents where actions and interactions follow from rules simulating a simplified representation of a real-world challenge or issue and various scenarios are tested Analysis of the data generated by running huge numbers of tests in this simulated environment provides insights into patterns that emerge from the introduction of a policy decision an innovation or other socio-economic technical political or environmental phenomenon

The structure of an agent-based model is visualised in Figure 7 As an example Table 1 displays these behavioural rules and subsequent (inter)actions as set out for a computer simulation of the Uganda rural water services system The first lsquothemersquo is the lsquowater service basis phasersquo which entails potentially replacing the current water user committee The second phase describes how the local government can be asked to assist in collecting money for the repair or maintenance of a water point The remaining four phases describe how the local government informs the district water officer about

State

Rules

Agent - based model

Modeller in the real world

Environment

State

RulesState

Rules

State

Rules

State

Rules

State

Rules

Agent

Actions en other agents

Actions en the Environment

FIGURE 7 STRUCTURE OF AN AGENT-BASED MODEL Source Nikolic et al 2013 p58

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 29: Change in complex adaptive systems

28

WORKING PAPER Change in complex adaptive systems

requests for a new water point or major repairs and the conditional grant proposal An agent-based model developed by Ghorbani (2013) combines the IAD framework of Ostrom (see Section 35) with theories from the field of sociology on the structure of human relations human agency and the influence of institutions on society to form an agent-based modelling framework through which socio-technical systems can be simulated A model created using the IAD framework is lsquoa diagnostic tool [which] starts from outcomesrsquo (Yu 2014 p56) Working backwards from one or more desired policy outcomes stakeholders can evaluate possible policy outcomes and revise policy objectives (Yu 2014 p56) Further they can lsquo identify patterns of interactionsrsquo and change the rules physical conditions actors and policies lsquoto understand how these variables affect the action arena and policy outcomesrsquo (Yu 2014 p56) By using Ghorbanirsquos modelling tool called Modelling Agents using Institutional Analysis researchers can analyse current policies or compare alternatives following a structured approach to decomposing the institutions (laws policies rules norms or shared strategies) that inform the behaviour of agents in the system For the purposes of this research project this tool offers a means for making explicit the requirements that the systemic change approaches discussed in Section 4 must satisfy to be formalised and simulated for further exploration and experimentation

The literature notes potential trade-offs between flexibility (the model can depict many types of behaviours with greater levels of ambiguity) and precision (the model accurately depicts the real-life system) (Miller and Page 2007 p79) Of critical importance is the process of model creation (Van Dam et al 2013) The effort to simulate a complex problem must engage domain experts and stakeholders in identifying the problem conceptualising the model and validating the model design so that the simulation generates useful results and insights for informing policy choices An issue that is not addressed in the literature about agent-based models is whether the results obtained by simulating and testing a range of future scenarios are easily understandable and actionable for policy makers politicians civil society representatives practitioners and other stakeholders involved in delivering public services The gap in the knowledge base about the acceptability and accessibility of agent-based modelling for decision makers involved in rural water services will be examined in the course of this research project

In Section 52 the literature about serious games is reviewed for its potential to bridge the gap between the the simulated results of agent-based models and the reality of decision-making processes for public services

52 SERIOUS GAMES

Serious games or workshop gaming approaches are lsquoa special type of model that uses gaming techniques to model and simulate a systemrsquo (Duke and Geurts 2004 cited in Duke and Kriz 2014 p145) With serious games concepts and technologies derived from computer-based and other entertainment games are used for non-entertainment purposes such as research policy development and analysis decision making training and learning

Serious games simulate the physical technical economic information communication and social elements of a real-world setting In this virtual environment actors can interact and engage in a competitive or non-competitive manner without the lsquorisk of real-world consequencesrsquo (Lukosch and Bekebrede 2014 p145) Gaining insight into the effect of large-scale changes before implementing them is of great value in policy making action research and programming in dynamic complex systems As noted by Chappin and Dijkema in Van Dam et al (2013 p217) a serious game about the European Unionrsquos energy market made participants lsquomore receptive to the fundamental complexity of the socio-technical electricity systemrsquo and helped them envision alternative scenarios (which were also enacted in serious games) before applying energy market trading strategies in the real world

A serious game may be as simple as pen and paper such as the participatory rapid appraisal methods common in the development sector But serious games may also be highly sophisticated high-fidelity computer-generated environments like those used to train medical professionals and aircraft pilots (Lukosch and Bekebrede 2014)

One serious game toolkit with decision support and didactic potential is Wat-A-Game developed by the Institut national de recherche en sciences et technologies pour l environnement et lrsquoagriculture (IRSTEA) It can be used to design participatory simulations (ie role-playing games) for water management policy design and education The game has been used in sub-Saharan African countries as well as in the Mediterranean region over the past decade to explore water-related issues by depicting water flows pollution resource sharing and uses

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 30: Change in complex adaptive systems

29

December 2015

Game participantsmdashfarmers household water users policy makers researchers implementersmdashdecide how they wish to use or allocate water resources as individuals or collectively and then observe the consequences for macro processes such as local economies user satisfaction labour and the environment After playing different scenarios groups of stakeholders engage in facilitated reflection and planning processes to inform the design of new water resource management policies

Just as for agent-based models and their results a question arises about the accessibility of serious games and their results and thus their utility to policy makers and practitioners The literature suggests that some combination of the two methods may have potentialmdashsomething to be explored in further research on achieving systemic change in the rural water sector

RulesTable 30 Rules of local government

Theme Action information Timing

Water service basis Check received lsquoinactive Water User Committeersquo information 1x day

Phase If False

Go to next phase

If True

Check if particular Water User Committee Motivation lt Motivation threshold

If False

Go to next phase

If True

Determine to replace Water User Committee or not

If False

Go to next phase

If True

Send Water User Committee replacement information

Monthly contribution Check received request money collection support 1x day

Phase If False

amp Request Go to next phase

Maintenance phase If True

amp Assessment phase Check water point User list

Send Obligation to pay Water User Committee the requested amount

Source Van Tongeren 2014 p185

TABLE 1 AGENT-BASED MODEL CONCEPTUALISATION RULES OF LOCAL GOVERNMENT

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 31: Change in complex adaptive systems

30

WORKING PAPER Change in complex adaptive systems

This review of literature has summarised findings from the domains of rural water services governance international development aid and its role in influencing local development processes and priorities as well as theory and concepts offered by the fields of systems thinking and complexity sciences The insights gained from this review include the validity of applying a complexity sciences perspective to framing the rural water sector as a systemmdashspecifically a socio-technical system

At the outset of this exercise in mid-2013 the initial line of inquiry was whether the complexity sciences literature had addressed whole system change as a phenomenon whether concerted approaches to fostering whole system change had been documented and whether the elements of such approaches had been identified or otherwise formalised The hope was that those elements could be compared or assessed for use in scaling up in other contexts

The review confirms that the notion of whole system change exists and has been evident in the literature from 1970s onwards The literature reviewed offers the concepts and theory from complexity sciences and in particular socio-technical systems as a more precise means for communicating about the rural water sector in low- and middle-income countries Framing this sector as a socio-technical system creates the possibility of examining how large-scale change may emerge and evolve in dynamic and complex systems

The literature also documents concerted approaches to achieving change in such a setting As applied in other public service contexts such as education health and water resources management approaches to whole system change include the following common elementsbull A commonly held vision of change among

stakeholders that has a broad base of supportbull Dedicated funded facilitation to guide and curate a

high-quality change processbull Evidence generated from monitoring policy and

action researchbull Collective or social learning and decision making

about alternative options and pathways to achieving the vision

bull Timely information feedback loops based on common monitoring systems and research across all levels of the system to keep stakeholders engaged and informed

The literature does not compare the different approaches or their elements however It is also not clear whether certain elements principles or strategic objectives are more critical or effective than other The literature indicates the value of competencies and methods that enable actors and actor networks to learn and reflect on what works and adapt to emerging change pressures in specific contexts but offers no set menu of options or principles that constitute a single approach

The literature does describe tools well suited to simulating and experimenting with complex problems for the purpose of identifying policy options and potential solutions These methods rooted in complexity sciences are powerful tools for exploring lsquoproblem spacesrsquo and complex global problems to use the terminology of Harman (1995)

The wider insights from this reviewmdashfor practical applicationmdashis another matter The Triple-S Sustainable Services That Last project which provided the context of this inquiry concluded its six-year programme cycle in November 2014 In addition to developing research reports tools for monitoring and policy analysis and development for sustainable delivery of rural water services Triple-S also sought to make explicit and to analyse its own programmatic approach to fostering change from an infrastructure-oriented approach to an emphasis on service delivery

Two Triple-S case studies document the approaches pursued by programme partners including government actors to create whole system change in Ghana [URL] and Uganda (forthcoming 2016) Additionally a series of blogs [URL] by Triple-S partners describe the implementation and testing of these approaches to systemic change Taken together these manuscriptsmdashincluding this literature reviewmdashdocument an evolution in the initiativersquos approach and its partnersrsquo collective thinking on how to achieve large-scale systemic change The goal of whole system change is now more explicitly articulated as the need for strong national systems (following Ryanrsquos definition of a system page 9) that have the capacity to act learn and adapt in the face of future uncertainties

Approaches to fostering whole system change exist and have been tested in different settings to overcome challenges and achieve a range of development goals Three approaches each grounded in complexity sciencesmdashcollective impact problem-

6 Discussion and conclusion

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 32: Change in complex adaptive systems

31

December 2015

driven iterative analysis and learning alliancesmdashare well suited to effecting change in socio-technical systems Particular aspects of these approaches such as the imperative of practitioners and policy makers to experiment and learn from different interventions and ideas are highlighted as strategic objectives and principles Involving a broad range of interested parties is crucial for attaining richer and more appropriate results as is willingness to adapt actions and plans according to emerging dynamics and signals

Furthermore the literature from the complexity sciences offers concepts and methods for exploratory inquiry into change processes in socio-technical systems The concepts from ADICO (Ostrom) enable the formalisation of the prevailing institutions in a given socio-technical system and simulation methods such as agent-based modelling provide a way to anticipate agentsrsquo behaviour and the emerging marco-level patterns Researchers can thus experiment with approaches to fostering systemic change over long periods of time and under varying conditionsmdashsomething not possible in real time

What the literature review has not revealed is which elements of the approaches to fostering systemic change are most effective Additionally no comparative study of the approachesrsquo strengths and weaknesses or even a framework for such a comparison was found Note however that this literature review predates the more recent articulation of the notion of national systems strengthening by the research team as the overarching aim of engaging with a whole system for improving public services

Next steps in the research programme will therefore focus on addressing gaps in knowledge about what is required to achieve change in national systems dedicated to delivering public services The three approaches discussed in Section 4 will be analysed and compared for their efficacy An additional line of inquiry will consider whether and how exploratory agent-based modelling and analysis can generate insights for policy makers and practitioners or whether other methods such as serious games would make results from agent-based simulations more accessible and useful

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 33: Change in complex adaptive systems

32

WORKING PAPER Change in complex adaptive systems

Aldrich HE Hodgson GM Hull DL Knudsen T Mokyr J Vanberg VJ 2008 In defence of generalized Darwinism Journal of Evolutionary Economics 18 pp577ndash96 doi101007s00191-008-0110-z

Andrews M Pritchett L and Woolcock M 2012 Escaping Capability Traps through Problem-Driven Iterative Adaptation (PDIA) CGD Working Paper 299 Washington DC Center for Global Development Available at httpwwwcgdevorgcontentpublicationsdetail1426292 [Accessed 1 September 2014]

Andrews M Pritchett L and Woolcock M 2013 Escaping capability traps through problem-driven iterative adaptation (PDIA) World Development 51 pp234ndash44 doi 101016jworlddev201305011

Bache I 2003 Governing through governance Education policy control under New Labour Political Studies 51(2) pp300-314

Barder O 2012 The implications of complexity for development Transcription by the author of O Barder Kapuscinski Development Lecture University of Bucharest May 15 2012 [online] Available at httpinternationalcgdevorgmediaimplications-complexity-development-owen-barder [Accessed on 30 August 2013]

Bar-Yam Y 1997 Dynamics of complex systems Boulder Colorado Westview PressBar-Yam Y 2005 Making things work Solving complex problems in a complex world Cambridge Massachusetts

NECSI Knowledge PressBlyth M Hodgson GM Lewis O and Steinmo S 2011 Introduction to the special issue on the evolution of

institutions Journal of Institutional Economics September 2011 7(3) pp299ndash315 doi101017S1744137411000270

Bramson RA and Buss T 2002 Methods for whole system change in public organizations and communities An overview of the issues Public Organization Review 2 pp211ndash21

Briggs X 2008 Democracy as Problem-Solving Civic Capacity in Communities Across the Globe Cambridge Massachusetts MIT Press

Brinkerhoff DW 2010 Review of Systemic action research A strategy for whole system change by Danny Burns Public Organization Review 10 pp93ndash95 doi101007s11115-009-0105-8

Burke M 2006 Robustness resilience and adaptability implications for national security safety and stability (Draft) Technical Report Defence Science and Technology Organisation

Burns D 2007 Systemic action research A strategy for whole system change Policy Press Bristol UK [online] Available at httpbooksgooglenlbookshl=enamplr=ampid=tx9pYZkwxugCampoi=fndamppg=PR4ampdq=systemic+action+research+strategy+whole+system+changeampots=qm3t4R_1Jjampsig=OumeZUct8WuIFRWGoNC6xHwwmUMv=onepageampq=systemic20action20research20strategy20whole20system20changeampf=false [Accessed on 29 October 2013]

Calvert R 1995 The rational choice theory of social institutions Cooperation coordination and communication In Banks JS and Hanushek EA eds 1995 Modern political economy Old topics new directions Cambridge University Press Cambridge Ch2 Pp216ndash67

Casella D Magara P Otim R and Watsisi M 2014 Triple-S Uganda Learning Facilitator Workshop Series 2013ndash2014 Unpublished report IRC The Hague and Kampala Uganda

Casella D Magara P Nabunnya JM Otim R Watsisi M Abisa J Mirembe L Lockwood H Moriarty P and Carriger S 2015 Whole system change capturing the change process in the Uganda rural water sub-sector Working Paper IRC The Hague

Connor M 2001 Developing network-based services in the NHS International Journal of Health Care Quality Assurance Incorporating Leadership in Health Services 14(6-7) pp 237-44

Connor M and Kissen G 2010 Tackling whole-systems change The Trafford framework for integrated services Journal of Integrated Care 18 pp4ndash14 doi105042jic20100243

Crawford SE and Ostrom E 1995 A grammar of institutions American Political Science Review 89(3) pp582ndash600

Dattee B and Barlow J 2010 Complexity and whole-system change programmes Journal of Health Services Research amp Policy 15 pp19ndash25 doi101258jhsrp2009009097

De Bruijn J A and E F Ten Heuvelhof 2003 Policy analysis and decision-making in a network How to improve the quality of analysis and the impact on decision-making Impact Assessment and Project Appraisal 20 pp 232ndash242

References

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 34: Change in complex adaptive systems

33

December 2015

De la Harpe J 2007 Strengthening local governance for improved water and sanitation services [online] IRC The Hague Available at httpwwwircwashorgresourcesstrengthening-local-governance-improved-water-and-sanitation-services [Accessed on 13 October 2014]

Da Silva Wells C 2012 Learning alliances from innovation to transformation [blog] 14 February 2012 Available at httpwwwircwashorgbloglearning-alliances-innovation-transformation [Accessed on 30 November 2014]

Darwin CR 1859 On the origin of species by means of natural selection or the preservation of favoured races in the struggle for life Murray London

Department for International Development (DFID) 2007 Governance development and democratic politics DFIDrsquos work in building more effective states Policy Booklet London Available at httpwebarchivenationalarchivesgovuk+httpwwwdfidgovukpubsfilesgovernancepdf [Accessed on 13 October 2014]

Dijkema GPJ Lukszo Z and Weijnen MPC 2013 Introduction In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch1

Douthwaite B Keatinge JDH and Park JR 2002 Learning selection An evolutionary model for understanding implementing and evaluating participatory technology development Agricultural Systems 72 pp109ndash31 doi 101016S0308-521X(01)00071-3

Duffy FM Reigeluth CM Solomon M Caine G Carr-Chellman AA Almeida L Frick T Thompson K Koh J Ryan CD and DeMars S 2006 The process of systemic change Step-up-to-excellence A protocol for navigating whole-system change in school districts TechTrends 50 pp41ndash51 doi101007s11528-006-7585-y

Duke R D and Kriz WC 2014 Back to the future of gaming W Bertelsmann Verlag BielefeldEdwards N Rowan M Marck P and Grinspun D 2011 Understanding whole systems change in health care

The case of nurse practitioners in Canada Policy Politics amp Nursing Practice 12 pp4ndash17 doi1011771527154411403816

Epstein JM and Axtell R 1996 Growing artificial societies social science from the bottom up Washington DC Brookings Institute

Flood RL 2010 The relationship of ldquosystems thinkingrdquo to action research Systemic Practice and Action Research 23(4) pp269ndash84 doi101007s11213-010-9169-1

Garandeau R Casella D and Bostoen K 2009 Evaluating and improving the WASH sector Strengthening WASH governance learning about complexity assessing change Thematic Overview Paper 23 The Hague

Gardner Roy and Elinor Ostrom 1991 Rules and Games Public Choice 70(2) May pp121-49Geels FW 2004 From sectoral systems of innovation to socio-technical systems Insights about dynamics and

change from sociology and institutional theory Research Policy 33(6-7) pp 897-920Ghorbani A Ligtvoet A Dijkema G and Nikolic I 2010 Using institutional frameworks to conceptualize

agent-based models of socio-technical systems In CoSMoS (Complex Systems Modelling and Simulation) 2010 Proceedings of Third Workshop on Complex Systems Modelling and Simulation Odense Denmark August 19 United Kingdom Luniver Press

Ghorbani A 2013 Structuring Socio-Technical Complexity Modelling agent systems using institutional analysis Doctoral thesis Delft University of Technology The Netherlands

Greenhalgh T Macfarlane F Barton-Sweeney C and Woodard F 2012 ldquoIf we build it will it stayrdquo A case study of the sustainability of whole-system change in London Milbank Quarterly 90(3) pp516ndash47 doi101111j1468-0009201200673x

Grindle M 2004 Good enough governance poverty reduction and reform in developing countries Governance An International Journal of Policy Administration and Institutions 17 pp525-48

Harman WW 1995 Global dilemmas and the plausibility of whole-system change Technological Forecasting and Social Change 49(1) pp1ndash12

Heifetz RA Kania JV and Kramer MR 2004 Leading Boldly Foundations can move past traditional approaches to create social change through imaginative ndash and even controversial ndash leadership (Stanford Social Innovation Review Winter 2004) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgimagesarticles2004WI_feature_heifetzpdf [Accessed on 7 February 7 2014]

Hodgson GM 2008 How Veblen generalised Darwinism Journal of Economic Issues 42(2) pp399ndash405Hodgson GM and Knudsen T 2010 Generative replication and the evolution of complexity Journal of

Economic Behaviour and Organisation 75(10) pp12-24 Available at httplinkinghubelseviercomretrievepiiS0167268110000454 [Accessed on 2 September 2013]

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 35: Change in complex adaptive systems

34

WORKING PAPER Change in complex adaptive systems

Holland JH 1992 Adaptation in natural and artificial systems an introductory analysis with applications to biology control and artificial intelligence Cambridge Massachusetts MIT Press

Huitema D Mostert E Egas W Moellenkamp S Pahl-Wostl C and Yalcin R 2009 Adaptive Water Governance Assessing the Institutional Prescriptions of Adaptive (Co)-Management from a Governance Perspective and Defining a Research Agenda Ecology and Society 14(1) 26 Available at httpwww ecologyandsocietyorgvol14iss1art26 [Accessed on 1 September 2015]

Iyer KNS 2002 Learning in strategic alliances an evolutionary perspective Academy of Marketing Science Review 10 pp1-16

Kahangire P 2012 Study to improve efficiency and effectiveness of learning and reflection processes in and between the different coordination platforms in the Uganda WASH sector Consultancy report Kampala Uganda

Kania J and Kramer M 2011 Collective Impact (Stanford Social Innovation Review Winter 2011) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentrycollective_impact [Accessed 7 February 2013]

mdashmdashmdash 2013 Embracing emergence how collective impact addresses complexity (Stanford Social Innovation Review Winter 2013) [pdf] Stanford Leland Stanford Jr University Available at httpssirorgarticlesentryembracing_emergence_how_collective_impact_addresses_complexity [Accessed 7 February 2013]

Keohane R O and E Ostrom eds 1995 Local commons and global interdependence heterogeneity and cooperation in two domains London Sage

Khanna T Gulati R and Nohira N 1998 The dynamics of learning alliances competition cooperation and relative scope Strategic Management Journal 19 pp193-210

Kooiman J 1993 Modern Governance New Government-Society Interactions London SageLewis O and Steinmo S 2010 Taking evolution seriously in political science Theory in Biosciences 129(2-3)

pp235ndash45 doi101007s12064-010-0097-5mdashmdashmdash 2012 How institutions evolve evolutionary theory and institutional change Polity 44(3) pp314ndash339

doi101057pol201210Lockwood H and Smits S 2011 Supporting rural water supply moving towards a service delivery approach

Warwickshire UK Practical Action Publishing Available at httpwwwircwashorgresourcessupporting-rural-water-supply-moving-towards-service-delivery-approach [Accessed 15 May 2014]

Lockwood H Duti V Nkum J Korboe D Carriger S 2015 Whole system change capturing the change process in the Ghana rural water sub-sector Working Paper IRC The Hague Available at httpwwwircwashorgresourceswhole-system-change-capturing-change-process-ghana-rural-water-sub-sector [Accessed September 1 2015]

Lukosch H and Bekebrede G 2014 The future of gaming Challenges for designing distributed games In RD Duke and WC Kriz eds Back to the future of gaming Bielefeld Germany

Lundy M 2004 Learning alliances with development partners a framework for out scaling research results In Pachico D ed 2004 Scaling Up and Out Achieving Widespread Impact Through Agricultural Research Cali Colombia International Center for Tropical Agriculture (CIAT)

Lundy M Gottret M and Ashby J 2005 Learning alliances an approach for building multistakeholder innovation systems Brief 8 Institutional Learning and Change Rome

Lyall C and Tait J 2004 Foresight in a multi-level governance structure Policy integration and communication Science and Public Policy 31(1) pp 27-37

Lyall C and Tait J 2005 New modes of governance Developing an integrated policy approach to science technology risk and the environment Aldershot UK Ashgate

Manning MR and De la Cerda J 2003 Building organizational change in an emerging economy Whole systems change using large group interventions in Mexico Research in Organizational Change and Development 14 pp51ndash97 [online] Available at httpwwwemeraldinsightcomdoipdf101016S0897-301628032914080-3 [Accessed 17 January 2014]

Miller JH and Page SE 2007 Complex adaptive systems an introduction to computational models of social life Princeton Studies in Complexity Princeton New Jersey Princeton University Press

Moriarty P Fonseca C Smits S and Schouten T 2005 Learning alliances for scaling up innovative approaches in the water and sanitation sector Conference background paper Learning Alliance Symposium UNESCO-IHE June 6ndash10 Delft The Netherlands IRC

Mowles C Stacey R and Griffin D 2008 What contribution can insights from the complexity sciences make to the theory and practice of development management Journal of International Development 20 pp 804-820

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 36: Change in complex adaptive systems

35

December 2015

Nikolic I 2009 Co-evolutionary method for modelling large scale socio-technical systems evolution Doctoral thesis Delft University of Technology The Netherlands

Nikolic I and Kasmire J 2013 Theory In KH van Dam et al eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch2

Nikolic I and Ghorbani A 2011 A method for developing agent-based models of socio-technical systems In Proceedings of the IEEE International Conference on Networking Sensing and Control April 11ndash13 2011 Delft The Netherlands doi101109ICNSC20115874914

Nikolic I van Dam KH and Kasmire J 2013 Practice In KH van Dam et al eds Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer Ch3

Nimanya C et al 2011 Uganda Lessons for rural water supply Assessing progress towards sustainable service delivery Triple-S Uganda country study The Hague and Kampala Uganda IRC International Water and Sanitation Centre

Nkum J et al 2014 Research on learning alliance approach Consultancy report Nkum Associates Accra Ghana

Organisation for Economic Co-operation and Development (OECD) 2009 Managing water for all An OECD perspective on pricing and financing key messages for policy makers Paris

Ostrom E 2011 Background on the institutional analysis and development framework Policy Studies Journal 39(1) pp7ndash27 doi101111j1541-0072201000394x

Ostrom E and Janssen M 2004 Multilevel governance and resilience of social-ecological systems In Spoor M ed 2004 Globalization Poverty and Conflict Dordrecht Kluwer Academic Publishers

Pahl-Wostl C 2009 A conceptual framework for analysing adaptive capacity and multi-level learning processes in resource governance regimes Global Environmental Change 19(3)(August) pp354ndash65

Pahl-Wostl C Becker G Knieper C Sendzimir J 2013 How multilevel societal learning processes facilitate transformative change a comparative case study analysis on flood management Ecology and Society 18(4) [online] Available at httpwwwecologyandsocietyorgvol18iss4art58 [Accessed 7 July 2014]

Pahl-Wostl C Sendzimir J Jeffrey P Aerts J Bergkamp G and Cross K 2007 Managing change toward adaptive water management through social learning Ecology and Society Special Feature on New Methods for Adaptive Water Management 12(2) [online] Available at httpwwwecologyandsocietyorgvol12iss2art30 [Accessed 7 July 2014]

Pierre J and Peters BG 2000 Governance politics and the state Basingstoke MacmillanPlummer J and Slaymaker T 2007 Rethinking governance in water services London Overseas Development

Institute Available at httpwwwodiorgsitesodiorgukfilesodi-assetspublications-opinion-files602pdf [Accessed 1 September 2015]

Polski M and Ostrom E 1999 An institutional framework for policy analysis and design Working paper Workshop in Political Theory and Policy Analysis Indiana University Bloomington

Prigogine I Stengers I and Toffler A 1984 Order out of Chaos Mans new dialogue with nature New York Bantam

Pritchett L Woolcock M and Andrews M 2010 Capability traps The mechanisms of persistent implementation failure Working paper 234 Washington DC Centre for Global Development

Pritchett L Samji S and Hammer J 2013a Its all about MeE Using structured experiential learning (ldquoerdquo) to crawl the design space Working Paper 322 Washington DC Center for Global Development Available at httpwwwcgdevorgsitesdefaultfilesits-all-about-mee_1pdf [Accessed 30 August 2014]

Pritchett L Woolcock M and Andrews M 2013b Looking like a state Techniques of persistent failure in state capability for implementation Journal of Development Studies Special Issue on State Society and Governance 49 pp1ndash18 doi101080002203882012709614

Ramalingam B 2013 Aid on the edge of chaos rethinking international cooperation in a complex world Oxford Oxford University Press

Rhodes RAW 1997 Understanding governance policy networks governance reflexivity and accountability Buckingham Open University Press

Rogers P and Hall AW 2003 Effective water governance Technical background paper 7 Stockholm Global Water Partnership

Rittel HWJ and Webber MM 1973 Dilemmas in a general theory of planning Policy Sciences 4 pp155ndash69 doi101007BF01405730

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 37: Change in complex adaptive systems

36

WORKING PAPER Change in complex adaptive systems

RWSN 2012 Rural Water Supply in the 21st Century Myths of the Past Visions for the Future Report of the 6th International Rural Water Supply Network Forum 2011 Uganda Kampala Rural Water Supply Network Available at httpwwwircwashorgsitesdefaultfilesRWSN-2011-Ruralpdf [Accessed 24 January 2014]

Ryan A 2008 What is a systems approach Available at httparxivorgpdf08091698pdf [Accessed 15 November 2015]

Schouten T and Moriarty P 2003 Community water community management From system to service in rural areas London Intermediate Technology Development Group Publishing (Practical Action)

Schouten T and Moriarty P 2013 The Triple-S Theory of Change Working Paper 3 The Hague IRC International Water and Sanitation Centre

mdashmdashmdash 2013 The Triple-S theory of change Working paper 3 httpwwwwaterservicesthatlastorgcontentdownload172818887version1file2013_WP3_Theory+of+Changepdf [Accessed 24 January 2014]

Sen A 1992 Inequality reexamined New York and Cambridge Massachusetts Russell Sage Foundation and Harvard University Press

Senge P 1990 The fifth discipline The art and practice of the learning organisation DoubledayCurrency New York

Smits S Moriarty P and Sijbesma C 2007 Learning alliances Scaling up innovations in water sanitation and hygiene Technical paper series no 47 Delft IRC International Water and Sanitation Centre

Smits S Verhoeven J Moriarty P Fonseca C and Lockwood H 2011 Arrangements and cost of providing support to rural water service providers Working paper 5 WASHCost The Hague IRC

Stoelhorst JW and Huizing A 2006 The firm as a Darwin machine an evolutionary view of organizational knowledge and learning The Academy of Management Annual Meeting Proceedings 2006 Atlanta

Stoker G (1998) Governance as theory Five propositions International Social Science Journal 50(155) pp 17-28Taylor B 2009 Addressing the sustainability crisis Lessons from research on managing rural water projects

WaterAid Tanzania [online] Available at httpwwwwateraidorg~mediaPublicationssustainability-crisis-rural-water-management-tanzaniapdf [Accessed 24 January 2014]

Trist E 1981 The evolution of socio-technical systems A conceptual framework and an action research program In AH Van de Ven and WF Joyce eds Perspectives on organization design and behavior New York Wiley Ch 2

UNICEFWHO 2010 Progress on Sanitation and Drinking-water 2010 Update WHOUNICEF Joint Monitoring Programme for Water Supply and Sanitation Geneva

United Nations 2009 World Urbanization Prospects The 2009 Revision United National Population Division Department of Economic and Social Affairs New York Available at httpwwwunorgendevelopmentdesapopulationpublicationspdfurbanizationurbanization-wallchart2009pdf [Accessed on 16 February 2015]

Van Dam KH Nikolic I and Lukszo Z eds 2013 Agent-based modelling of socio-technical systems Dordrecht The Netherlands Springer

Van Tongeren SAE 2014 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Applied to the water services delivery system in rural areas of Uganda in an agent-based model design MSc thesis final report Delft University of Technology The Netherlands

Van Tongeren SAE 2013 Creating a conceptual framework for a deeper understanding of evolving processes in socio-technical systems Draft Delft University of Technology The Netherlands

Veblen T 1899 The theory of the leisure class An economic study of institutions New York MacMillanmdashmdashmdash 1898 Why is economics not an evolutionary science Quarterly Journal of Economics 12(4) pp373ndash97Waldrop MM 1992 Complexity the emerging science at the edge of order and chaos New York Simon amp

SchusterWaterAid 2011 Off-track off-target Why investment in water sanitation and hygiene is not reaching those who

need it most Policy report London WaterAidWhite L 2000 Changing the ldquowhole systemrdquo in the public sector Journal of Organisational Change

Management 13 pp162ndash77 doi10110809534810010321481Williamson OE 2000 The new institutional economics Taking stock looking ahead Journal of Economic

Literature 38 pp595ndash613

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 38: Change in complex adaptive systems

37

December 2015

Woolcock M 2013 Using case studies to explore the external validity of ldquocomplexrdquo development interventions WIDER working paper 2013096 United Nations World Institute for Development Economics Research Helsinki Available at httpwwwwiderunuedupublicationsworking-papers2013en_GBwp2013-096_files90513411817668621defaultWP2013-096pdf [Accessed 18 October 2013]

mdashmdashmdash 2014 The science of delivery and the art and politics of institutional change [ppt] Lecture at Overseas Development Institute London UK

World Health Organisation (WHO) 2012 UN-water global analysis and assessment of sanitation and drinking-water (GLAAS) 2012 report The challenge of extending and sustaining services Geneva Available at httpwwwunorgwaterforlifedecadepdfglaas_report_2012_engpdf [Accessed 13 October 2014]

Yu C 2014 Eco-transformation of industrial parks in China Doctoral thesis Delft University of Technology The Netherlands

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg

Page 39: Change in complex adaptive systems

Visiting addressBezuidenhoutseweg 22594 AV The HagueThe Netherlands

Postal addressPO Box 823272508 EH The HagueThe Netherlands

T +31 70 3044000infoircwashorgwwwircwashorg